Intelligent Energy Management in Hybrid Electric Vehicles

Intelligent Energy Management in Hybrid Electric Vehicles 147
8
X
Intelligent Energy Management

in Hybrid Electric Vehicles
Hamid Khayyam1, Abbas Kouzani1, Saeid Nahavandi1,
Vincenzo Marano2 and Giorgio Rizzoni2
Deakin University Australia1 and The Ohio State University USA2
1. Introduction
Energy management in vehicles is an important issue because it can significantly influence the
performances of the vehicles. Improving energy management in vehicles can deliver
important benefits such as reducing fuel consumption, decreasing emission, lower running
cost, reducing noise pollution, and improving driving performance and ease of use. According
to Mainins (Manins, 2000), each year more then 50 million new cars are produced in the world.
However, usually only 30% to 40% of the energy produced by the engine is used to drive a car.
The large energy waste of around 60% is the result of having an engine powerful enough to
cope with the maximum power demand despite the fact that such power is required for only a
vary small percentage of vehicles’ operating time. In addition, vehicle emissions are a source of
greenhouse gas pollution emitting 70% to 90% of urban air pollution (SOE, 2006). Fuel
economy benchmarks and emission regulations have encouraged vehicle manufactures and
researchers to investigate new technologies to enhance fuel economy and minimise emissions.
The energy efficiency of vehicles can be improved by enhancing the efficiency of the vehicle.
Implementing energy management strategies in classical vehicles does not fully deliver the
expected benefits. Hybrid electric vehicles, on the other hand, offer a platform that can
accommodate advanced energy management strategies giving rise to full realization of the
stated benefits. Intelligent energy management methods can observe and learn driver
behavior, environmental and vehicle conditions, and intelligently control the operation of
the hybrid electric vehicle.
A Hybrid Electric Vehicle (HEV) takes advantage of an Internal Combustion Engine (ICE)
and an Electric Motor (EM) to deliver fuel consumption and exhaust emission reduction. An
EM is powered by on-board battery packs to drive the vehicle. From the consumers overall
perspective, the HEV is essentially the same as a Conventional Vehicle (CV). Moreover,
HEVs are refuelled in the same way as a CV. A HEV has the advantage over a pure Electric
Vehicle (EV) in both travelling range and convenience, as there is no need to recharge the
battery through a power point for long hours. Importantly, a HEV has the potential to
improve fuel economy by almost 50%, while also possessing all the advantages and
flexibility of a CV (Ehsani et al,. 2005). Hence, HEVs solve the problems of EVs whilst
minimising the shortcoming of CVs providing the benefits of both electric and conventional
www.intechopen.com
148 Energy Management
vehicles. HEVs are categorised into three groups: Series (S-HEV), Parallel (P-HEV), and
Series/Parallel (S/P-HEV) as shown in Fig. 1.
In an S-HEV, there is no mechanical link between the ICE and drive train. This means that
the ICE can run continuously in its preferred operating range, whereas the drivetrain is
driven by an electric machine. For the electric power request, it relies on the battery plus the
generator. The generator is driven by the ICE and maintains an appropriate energy level in
the battery. A disadvantage of this configuration is that energy is first converted from
mechanical power to electric power with the generator and then back to mechanical power
by the electric machine, both introducing losses.
The P-HEV establishes a parallel connection between the ICE and the electric machine that
both are allowed to give force to the drive the vehicle. The power through the EM can be
positive as well as negative. This allows the EM to operate in motor mode and generator
mode. At a top-level view, the P-HEV configuration looks similar to a conventional vehicle,
although the EM in a conventional vehicle operates only in generator mode.
Finally, the last vehicle configuration is an S/P-HEV. It merges the topology of a series and a
parallel HEV. S/P-HEVs have the highest complexity since power to the drivetrain can follow
various trajectories. Recently plug-in hybrid electric vehicle (PHEV) has come to market. A
PHEV is a hybrid electric vehicle that described above. The PHEV batteries can be recharged
by plugging into an electric power source. A PHEV combines type of conventional hybrid
electric vehicles and battery electric vehicles, possessing both an internal combustion engine
and batteries for power. The desire strategy using PHEV can be employed as follows: in short
distance travelling electric vehicle (EV) mode operation such as urban and for long distance
travelling hybrid electric vehicle (HEV) mode operation such as highways.
The most important challenge for the development of P-HEV is the synchronization of
multiple energy sources and conversion of power flow control for both the mechanical and
electrical paths in optimal fuel efficiency and battery areas.The difficulty in the development
of hybrid electric vehicles is the coordination of multiple sources such as mechanical and
electrical. The reason why a P-HEV is considered in this work is that it has fewer disadvantages
and less complexity (Kessels.J, 2007) (Ehsani et al.,2005).
Fig. 1. Three HEVs structures.
Nevertheless, any vehicle needs to deal with uncertain factors such as environment
conditions and also driver behaviour. HEVs are a highly complex systems comprising a
www.intechopen.com
large number of mechanical, electronic, and electromechanical elements (Zhu et al.,2002).

Hence a HEV can be considered as a Complex System (CS).
A Complex System is a system that can be analyzed into many components having
relatively many relations among them, so that the behaviour of each component depends on
the behaviour of others (Simon. A.H,1973).
In the real world, many problems and systems exist that are too complex or uncertain to be
represented by complete and accurate mathematical models. However, such systems need
to be designed, optimized, and controlled. CSs can be handled by Intelligent Systems (ISs).
ISs can learn from examples, are fault tolerant, are able to deal with non-linear problems,
and once trained can perform prediction and generalization at high speed. Intelligence
systems have been used in diverse applications in control, robotics, pattern recognition,
forecasting, medicine, power systems, manufacturing, optimization, signal processing, and
social/psychological sciences. They are useful in system modelling such as in implementing
complex mappings and system identification. ISs comprise areas like expert systems,
artificial neural networks, genetic algorithms, fuzzy logic and various hybrid systems,
which combine two or more of these techniques. ISs play an important role in modelling
and prediction of the performance and control of energy and renewable energy processes.
According to literature, ISs have been applied to energy and renewable energy engineering.
ISs can be developed through modelling and simulation. The modelling and simulation
approach has become an essential tool for mechanical engineers and automotive researches in
improving efficiency and timing of vehicle design and development, resulting in the delivery of
significant cost saving as well as environmental benefits. The modelling and simulation is
generally defined as mathematical realisation and computerised analysis of abstract
representation of systems. The modelling and simulation helps achieve insight into the
functionality of the modelled systems, and investigate the systems' behaviours and
performances. The modelling and simulation is used in a variety of practical contexts relating to
the design, development, and use of conventional as well as advanced vehicles including: design
and evaluation of vehicle performance, fuel consumption, emission, energy storage devices,
internal combustion engine, hybrid engine, accessories, composite materials, determination of
drag using wind tunnel, training drivers trough virtual vehicle, collecting and analysing sensory
information, identifying critical test conditions, investigating crash factors, characterising road
topology, testing and analysing energy management strategies, and so on.
This work employs the modelling and simulation approach to develop an Intelligent Energy
Management System (IEMS) for a P-HEV.
The main objective is to optimize fuel consumption and reduce emissions. The work involves the
analysis of the role of drivetrain, energy management control strategy and the associated impacts
on the fuel consumption with combined wind/drag, slope, rolling, and accessories loads.
2. Literature Review and Background

This section provides a review of the main approaches used in modelling and control of
energy management of HEVs. In a CV, energy can be dissipated in a number of ways
including (Kessels.J, 2007):
i. Brake utilisation: The brake is applied by the driver to decelerate the vehicle
resulting in the loss of kinetic energy in the form of heat.
www.intechopen.com
ii. Engine start/stop: The engine often runs idle during the utilisation of vehicle
resulting in an unnecessary consumption of fuel.
iii. Uneconomic engine operating condition: An engine often demonstrates nonlinear
fuel consumption behaviour in certain operating conditions that causes an
excessive use of fuel.
iv. Unscheduled load: Certain mechanical and electrical loads get activated outside the
economic operating point of engine increasing the fuel consumption.
P-HEVs provide a platform to reduce the wasted energy. The most important challenge for the
development of P-HEV is the synchronization of multiple energy sources and conversion of
power flow control for both the mechanical and electrical paths. Control in HEVs is recognized
as two levels of actions: supervisory control and component control. In this study supervisory
control is investigated as a suitable control strategy in energy management.
The control strategy is an algorithm that is used for issuing a sequence of instructions from
the vehicle central controller to operate the drivetrain of the vehicle. The control strategy
needs to monitor uncertain events. Moreover, in order to improve the system, the control
strategy can provide optimized energy management. The control strategies in a P-HEV can
be classified in two main groups as follows.
2.1 Rule-Based Control

The control rules techniques are based on mathematical, heuristics, and human expertise
generally with an analytical knowledge of a predefined driving cycle. Control rules can be
categorized in three methods.
A. Rule-Based
This method is based on an examination of the power requirements, ICE efficiency, fuel or
emission maps. Human knowledge is used to design rules to split the requested power between
converters. The method can be categorized into three groups: on/off control (Ehsani et al.,2005),
base line control (Zhu et al.,2002) (Sciarretta et al., 2004) (Linl et al.,2004) (Lyshevski,1999)
(Barsali et al., 2004) (Khayyam et al., 2008), and discrete time events (Zhang & Chen, 2001)
approaches.
B. Fuzzy logic
Fuzzy logic control has a nonlinear structure that can deal with the nonlinear structure of
the power split problem. Fuzzy logic has a more robust structure and offers more design
flexibility. The problem with fuzzy logic is the optimization and mathematical manipulation
of defuzzification system. The defuzzification process consumes memory and time in
controller. Some fuzzy logic controller have been developed for HEVs including (Baumann
et al., 2000) (Farrokhi & Mohebbi, 2005) (Langari &won, 2005) (Mohebbi et al.,2005) ( Salman
et al., 2000) (Schouten et al.,2002) (Hajimiri at al., 2008).
C. Neuro-Fuzzy
There are also combinations of fuzzy logic and artificial neural called neuro-fuzzy control
(Mohebbi et al.,2005) and fuzzy discrete event control (Bathaee et al., 2005).
www.intechopen.com
2.2 Optimal Control

In optimal control the controller is optimized according to a cost function of the system.
Therefore, optimal control strategies are almost perfect. However, the optimal controllers
are sensitive to parameter changes and also to noise. To perform the optimization process,
all the dynamic and static behaviours of the system components are taken into
consideration. Calculations are usually simplified by introducing assumptions which means
that the solution is optimum only under the assumptions. On the other hand, the discrete
time events method is simple and more robust. System behaviours are divided into discrete
events. Each event is connected to another by certain rules (Mohebbi & Farrokhi, 2007).
If this optimal control is performed over a fixed driving cycle, a global optimum solution
can be found. In fact, the optimal control system solution is noncasual in that it achieves the
reduction of fuel consumption using information of future and past power demands.
Obviously, this technique cannot be used directly for real-time energy management.
Optimal control can be divided in two groups as follows.
A. Global Optimization (off line)
There are several reported solutions to achieve performance targets by optimization of a cost
function representing efficiency over a drive cycle, yielding global optimal operating points.
The global optimization techniques are not directly applicable for real-time problems,
considering the fact that they are casual solutions. This is due to their computational
complexity. Some of the global optimization methods are given below:
A.1 Neural Networks

Neural networks have the ability to be trained online or offline, but online training
consumes memory in a controller. This trainability characteristic makes neural networks as
a good candidate for adaptive energy management systems. As an example, the work
presented in (Mohebbi &Farrokhi,2007) developed a neural network for optimal control.
Prokhorov (Prokhorov D.V , 2008) used a neural network controller for improved fuel
efficiency of the Toyota Prius hybrid electric vehicle. A new method to detect and mitigate a
battery fault was also presented. The developed approach was based on recurrent networks
and included the extended Kalman filter.
A.2 Classical Optimal Control

(Delprat et al,. 2004) used the optimal control theory based on Lewis and Syrmos (Lewis &
Syrmos., 1995) work. This method is directly applied to find a global solution for the energy
management problem in a parallel torque-addition arrangement. The analytical nature of
this method makes it a good one. However, variation of drivetrain structure makes it
difficult to find an analytical solution, compared with numerical and iterative-based
methods. Some optimal control have been developed for HEVs including (Wei et al.,2007)
(Pisu & Rizzoni,2007) (Musardo et al.,2007).
A.3 Linear Programming

This method can formulate the problem of optimizing the fuel efficiency as a nonlinear
convex optimization problem that is approximated by a large linear program (Tate &
www.intechopen.com
Boyd,1998). The approximations used for transformations and the fact that LP may not be
applicable to a more sophisticated drivetrain degrade the proposed approach.
A.4 Dynamic and Stochastic Programming

Dynamic Programming (DP) method utilizes the minimizing cost function over a driving
cycle. (Lin et al.,2003) demonstrated that the approach does not give a real-time solution by
nature. A family of random driving cycles needs to be used to find an optimal solution.
A.5 Genetic Algorithm

The Genetic Algorithm (GA) has been used to solve a constrained nonlinear programming
problem. (Piccolo et al.,2001) showed that GA is very useful for complex nonlinear optimization
problems. This is because GA leads to a more accurate exploration of the solution space than a
conventional gradient-based procedure. But GA dose not give the necessary view to the designer
of the powertrain , unlike an analytical approach. Montazeri et al. (Montazeri et al., 2006)
described the application of genetic algorithm for optimization of control parameters in P-HEV.
B. Real Time Optimization (on line)

In order to develop a cost function for real-time optimization, the following methods can be used.
B.1 Model Predictive Control

(Salman et al., 2005) utilized a look-ahead window to find a real-time predictive optimal
control law. This approach can be used for superior fuel economy by previewing the driving
pattern and road information.
B.2 Decoupling Control

(Barbarisi et al.,2005) proposed a novel strategy to assure acceptable drivability of the
vehicle that was based on the vehicle’s dynamic model. Based on the proposed decoupling
methods, the controller’s output is composed of different components.
B.3 Genetic-Fuzzy
The genetic-fuzzy control strategy is a fuzzy logic controller that is tuned by a genetic
algorithm. Poursamad et al. (Poursamad et al.,2008) and Montazeri et al. (Montazeri et al.,
2008) applied these control strategy model to minimize fuel consumption and emission.
2.3 Discussions
The presented work is focused on a control strategy to reduce fuel consumption though
considering performance and driveability. Our optimal control strategy is found in two
steps, first finding the control which results in the reduction of fuel consumption together
and offering the best performance, and second taking vehicle driveability into consideration.
Among the control strategies for the best fuel economy, dynamic programming is the only
one that assures global optimality if the driving cycle is known in advance.
However, it does not apply to real-time problems. On the other hand, fuzzy logic, rule-
based, and neuro-fuzzy controllers are not generally optimized, but applied to real-time
www.intechopen.com
problems. If the future driving conditions of a few minutes ahead can be predicted then the
optimal controller can help find a suboptimal solution.
3. Factors Involved in Energy Management of Hybrid Electric Vehicles

Bandivadekar and Heywood (Bandivadekar & Heywood, 2007) presented an analysis that
shows the possibility of halving the fuel consumption of new vehicles by 2035. Enhancement in
vehicle control and management strategies is considered to be an influential mean in reducing
the fuel consumption of vehicles. Energy management approaches in vehicles can be realised
through considering a number of factors including (Cacciabue et al., 2009): environmental
conditions, driver behaviour, vehicle specifications, and intelligent transportation approach
(EPA, 2004). In order to develop an energy management system, a number of models need to be
implemented and used. These models are described in the following.
3.1 Main factors involved in energy management system of HEVs

A HEV can be considered as a complex system consisting of subsystems. In the
development of energy management systems, model of the HEV subsystems are developed
and used. Fig. 2 shows an overview of the energy management model for HEVs.
Fig. 2. Overview of the energy management model for HEVs.
3.2 Model of Environment

Among the factors that are involved in HEV systems, the environment conditions such as road
geometrical specifications and wind behaviour are often unknown and uncertain during
drives. The information about the geometrical specification and wind behaviour of the road
www.intechopen.com
ahead of the vehicle can be used by an intelligent system to reduce fuel consumption of the
vehicles (Khayyam et al., 2008). However, this information is often unavailable to the
intelligent system on-board of a vehicle in real-time. Thus, utilising on-line and off-line
prediction and monitoring of the geometrical specifications and wind behaviour of the road
ahead of vehicles can improve their performances. Environmental information can be
categorized in two groups: current and look-ahead. The data include road geometry, road
friction, wind drag, and ambient temperature. It has been demonstrated that lookahead
environment information can be employed by the energy management system to achieve
reduction of fuel consumption (Hellstrom et al.,2009). Khayyam et al. (Khayyam et al.,2008)
presented a Slope Prediction Unit (SPU) to calculate the slope angle of the road within the
distance of 50-300 meters away from the vehicle. This information reduced fuel consumption
about 6.1% liter/100 km during simulation. Global Positioning Systems (GPS) and Geographic
Information Systems (GIS) can provide static and dynamic road information.
Current Environment Model (CEM) is an algorithm that creates data associated with
environmental conditions and frictions. Look-ahead Environment model (LEM) is an
algorithm that creates data associated with future environmental conditions and frictions
encountered by the vehicle.
In order to model environment, a number of methods can be used. Khayyam et al.
(Khayyam et al.,2009a) proposed a method that can be used to produce authentic highway
height data using a set of probability distributions. They considered a highway as a complex
road which can have any kind of possible geometrical variations. The presented method
models highway heights by Rayleigh probabilistic distribution function. In addition,
highway geometric design laws were employed to modify the created highway data making
it consistent with the real highway situation. The proposed model is then used to produce a
3D realistic road. The method is called a Probabilistic Highway Modelling (PMH) technique.
PMH is capable of creating artificial highway and wind data that possess statistical
characteristics of real highway and wind situations. A highway is considered to contain a
collection of road segments. The Poisson Probability Distribution Function (PDF) is used to
produce a random number that determines the number of road segments. Segments can
then have different lengths. For each segment, the exponential PDF produces a random
number that represents the segment length. In addition, for each segment, two other
random numbers are generated and used to form the geometry of the segment. The
Rayleigh PDF is employed to produce a random number that represents the height change
of the segment. Also, the Gaussian PDF is used to form a random number that gives the
bend deflection change in the segment. The random numbers for height and bend could be
small or large injecting varying degrees of heights and bends into different road segments.
Also, highway geometric design laws are used to modify the created highway data to make
it consistent with the real physical highway situation.
A wind is constructed using a collection of regions of differing lengths. A wind creation
algorithm is an iterative routine. The algorithm creates wind speed and direction values for
each region. The exponential PDF produces a random number that represents the region
length. The Weibull PDF is employed to produce a random number that represents the wind
speed value in the region. Also, the uniform PDF is used to form a random number that
gives the wind direction value in the region.
The PHM can be employed in simulation of problems involving highway roads such as
energy optimization of conventional and hybrid electric vehicles. Fig. 3 displays a flowchart
www.intechopen.com
diagram description of the highway creation algorithm using the PHM. The result of the
highway creation algorithm demonstrates in Fig. 4 that show a 3D representation of the
constructed sample highway using the PHM. Fig. 5 displays a flowchart diagram
description of the wind creation algorithm using the developed PHM concept.


Fig. 3. Highway creation algorithm using the developed PHM.
300
A ltitud e(m )
250
200
150
100
-1
-0.5 10
8 9
0 7
5 6
0.5 3 4
1 2
1 0
5 4
x 10 x 10
Y position(m)
X position(m)
Fig. 4. 3D representation of the constructed sample highway using PMH technique.
www.intechopen.com
Fig. 5. Wind creation algorithm using the developed PHM.
3.3 Model of Driver

Driver behaviour has a strong influence on emissions and fuel consumption of the vehicle.
Modelling driver behaviour can be done using different methods. As an example, the
Driver-Vehicle-Environment (DVE) (Cacciabue, 2007) (Lin et al., 2005) method models
human machine interaction and associated taxonomies for classifying human behaviour. De.
Vlieger et al. (De. Vlieger et al.,2000) identified three types of driving behaviour as follows:
1- Calm driving that implies anticipating other road user's movement, traffic lights,
speed limits, and avoiding hard acceleration.
2- Normal driving that implies moderate acceleration and braking.
3-Aggressive driving that implies sudden acceleration and heavy braking.
Moreover, they note that emissions obtained from aggressive driving in urban and rural
traffic are much higher than those obtained from normal driving. A similar trend is
observed in relation to fuel consumption. It is stated that the driving style affects the
emission rate and the fuel consumption rate.
Average acceleration and Standard Deviation (SD) of acceleration over a specific driving
range are used to identify the driving style. Acceleration criteria for the classification of the
www.intechopen.com
driver's style are based on the acceleration ranges proposed by De Vlieger et al. (De Vlieger
et al. 2000). They defined the typical ranges of average accelerations as describe in table 1.
Acceleration Calm Normal Aggressive

Driving Driving Driving
City 4.85-6.9 6.98-8.6 9.15-11.8
Journey(m/s2)
Highway 0.85 1.0 2.16
Journey(m/s2)
Table 1. Overview of the tested acceleration (De Vlieger et al., 2000).
Our objective is to use support dynamic real-time driver behaviour system in the energy
management system. A driver first determines the drive strategy, selects the engine
specifications, starts the vehicle motion, and controls the mass flow rate of fuel into ICE by
changing the pedal, gear, brake, and clutch. Also, the driver sends this data as drive strategy
to IEMS.
3.4 Model of Vehicle (Quasi-Static)

In a P-HEV, both the Internal Combustion Engine (ICE) and the Integrated
Starter/Generator (ISG) can give tractive force to the wheels. Furthermore, the ISG will be
used as a generator to supply the electric loads. A schematic drawing of the vehicle
configuration is shown in Fig. 6.
Es
Fig. 6. P-HEV topology (Kessels.J,2007).
the transmission and the clutch. The engine speed  and the drivetrain torque td are
The power demand of the drivetrain Pd covers all the elements of the drivetrain, including
calculated back from the vehicle speed and denote the driver’s power demand:
Pd   d (1)
The power split device is assumed to have no energy losses and establishes the following
power balance:
Pe = Pd + Phev (2)
Where : Phev is hybrid power and Pe is engine power.
www.intechopen.com
3.4.1 Conventional Vehicle Specification

A vehicle ICE can be treated as a controlled volume system whose energy balance is given
as follows:
Q combaustion  (Q fuel  Q air  Q exhaust ) (3)

 Proad friction  Pdrag  Pslope  Paccessory  Pdriving  Q water / oil  Q heatloss
In order to include all losses, Equation (3) is reformed into the following equation where the
effect of different losses is taken into account by corresponding efficiencies:
( Q combustion )   otto   fuel  air

  mechanical   heat loss  Pnet (4)
 Proad  friction  Pdrag  P slope  Paccessory  Pdriving
where: Pnet = Power output of engine
otto =Otto cycle efficiency=

1 
1 =0.529 (Pullkrabek,1997)
rc (   1 )
 = Real fuel air engine efficiency =0.75(Yaodong Wang,2007)

fuel  air
 =Mechanical efficiency=0.9 (Plint ,1997)

mechanical
 = Heat loss efficiency=0.8 (Pullkrabek,1997)

heat loss
These efficiency are depend on some variable factors and situations.They can be measured
by industrial vehicle companies. In the section 3.4.3 we will select specific efficiency in our
model.To calculate Q combaustion Equation (3) is used:
Q combustion  m fuel  qcombustion (5)
where qcombustion is the combustion energy. In this model, the fuel is assumed to be Cn H1.87 n in
(Wang et al., 2007) . The complete combustion of C8 H 14.96 with 1+k percent theoretical air is
written as:
C 8 H 14.96  k  11 .74  (O2  3.76 ) N 2 

 (6)
8CO 2  7.48 H 2 O  11 .74 ( k  1)O2  11 .74  3.76  k  N 2
If the heat transfer was accurately measured, the released energy would be 109100 kJ per
kg
8 mole of CO2 (Heywood .B.J, 1998). The result of Equation (6) gives:
q combustion   ni [ hf   h ] i  W C .V .   n e [ h f   h ]e (7)

Where:
n i [ h f   h ] i  h f C 8 H 14 . 96  1 . 2  11 . 74 ( O 2  3 . 76  N 2 )  793 . 23 kJ / kg
www.intechopen.com
Type Symbol Value

Description
Combustion
Enthalpy of formation E-F  Thermodynamic tables
h
 h
f
Sensible formation E-F Thermodynamic tables

Combustion energy E-F qcombustion 38017 kJ
kg
Mass flow rate of fuel V-O . kg
combustion m fuel s
Temperature of fuel V-S T fuel 27 °C

Temperature of air V-S T a ir 27 °C
Temperature of exhaust V-S Texhaust 450 °C
Engine compression ratio E-F rc 8.6
Air compression ratio E-F  1.35
Ratio of nitrogen per oxygen E-F rN 2 / O 2 3.76
Excess air V-O E excessair 20%
Road
Road friction E-F F friction C rolling mg cos 
Road friction coefficient E-F Crolling 0. 01
Gravity acceleration E-F g 9.8 m/s2
Vehicle velocity V-O V1 16.6 m/s
Vehicle angle V-O  1 0°
Drag
Drag friction E-F ( )  V A ( )
1
Fdrag C drag
2

2
Wind angle of attack E-F 2 Random direction (0-360o)
Wind velocity E-F V2 0-6 m/s
Result wind and vehicle angle V-O  Calculate in simulation IEMS
Result of wind and vehicle speed V-O Vt Calculate in simulation IEMS
 ( 0 . 00005 )     0 . 0097     0 . 31
Result of wind and vehicle speed V-O Vt 1 Calculate in simulation IEMS
Drag coefficient (By simulation ) V-S Cdrag ( )
Front surface area V-S A( ) 1.8 1 cos 
m

Vehicle + passenger mass V-O 1280 kg
Air density E-F 1.225 kg/m3
Slope
Slope friction R-O Fslope mg sin 
Road slope angle R-O  – 1% ≤ atan(Ø) ≤ +0.6%
Radius of Comfort requirement R-O R 100 m
Accessory
Accessory V-O Paccessory 0-4250 watt
V-S vehicle specification; V-O vehicle operation; E-F environment factors;
R-O road condition.
Table 2. Parameters involved in energy balance equation
www.intechopen.com

and
n e [ h f  h ]e  8CO 2 [ h f   h ]CO 2  7.48 H 2 O[ h f   h ]H 2 O  11 .74  (0 .2)O 2 [  h ]O 2 
11 .74  3.76  1.2  N 2 [  h ] N 2  37219 .70 kJ / kg
and WC .V =0
q combustion  |  37219 . 70  793 . 23 | 38017 . 93 kJ / kg (8)
Substituting the terms stated in Table 2, the mass flow rate fuel consumption of the vehicle
Proad friction  Pdrag  Pslope  Paccessory  Pdriving

can be calculated as follows :
 fuel 
(9)
(qcombustion) otto  fuelair mechanicalheat loss
m
The total fuel consumption in this process is:
  m  dt
T
m (10)
fu e l fu e l
0
Pnet  [ F friction  F drag  F slope  F ] Vt  [1 / 2  m  (V t 2 1  V t )]

1
t
2
accessory
where t is the total numbers of steps involved in the simulation.
The symbols given in these equations are described in Table 2. The acceleration of the
vehicle in t time can be calculated as:
V t  V t 1 dV
at   (11)
t dt
Also, the distance traversed by vehicle in t is:
Xt   a t   t 2  V t 1   t
1
2 (12)
3.4.2 Parallel Hybrid Electric Vehicle Specification

The ISG is mounted on the crankshaft of the ICE and therefore, it is also coupled to the drive
train of the vehicle. Since the ISG model uses power based signals, it is not possible to
observe speed-dependent characteristics. The ISG operates similar to the electric machine. It
can operate in two modes: generator mode (Phev>=0 ) and motor mode ( Phev<0).
The electric power net connects the ISG with the electric loads and the battery. No losses are
assumed in the electrical wires, leaving the following description:
Pc  Pem  Pb (13)
Where: Pem is electric machine power, Pb battery power, and Pc electric loads.
The battery model consists of two subsystems: a static efficiency block and a dynamic
energy storage block, see Fig. 7. The battery model is used where the losses grow
proportionally with the power during charging (Pb > 0) and discharging (Pb < 0).
www.intechopen.com
The efficiency block incorporates the energy losses during charging and discharging,
whereas the energy storage block keeps track of the actual energy level Es in the battery. At
this point an integrator is used:
E s (t e )  E s (0 )  P ( t ) dt , Pb  max(   Ps , 1 Ps )
t

s
(14)
0
Es  bat
Fig. 7. Battery Model
To indicate the actual charging level of the battery, the State of Charge (SOC) is often used.
However, the physical background of SOC has a strong relation with battery models based
on current and voltage. Because the proposed battery model is power based, the State of
Energy (SOE) is more appropriate. The SOE expresses the relative energy status as follows:
SOE   100 %
Es
(15)
E cap
Depending on the control strategy from the EM system, three different representations of
the internal battery losses are taken into account, which approximate the relation between
the power Pb at the battery terminals and the net internal power Ps. Table 3 provides the
specifications of the battery and EM. The battery efficiency is considered as:
 Batt  2547600
2881008
 88 % (16)
Feature Symbo Type

l
Battery (NHW11)
Cells per module 6
Total Volts Vmax 273.6
Capacity (Amp hours) 6.5
Capacity (Watt hours) 1778.4
Electric Motor
Operating Voltage (V) Vmin 273
Power (W) 33000-44000
Table 3. Parameters involved in energy balance equation
3.4.3 Control strategy and optimal torque

The control strategy involves calculation of the torque produced by ICE based on various
parameters such as road load and battery SOC. This includes the calculation of an optimal
torque based on contending ICE parameters, and deciding the actual torque output by later
modifying the optimal torque based on road load and battery SOC. The optimal torque map
is shown in Fig. 8.
www.intechopen.com
Fig. 8. Optimal torque map
At the same current speed, if the required torque is above the optimal torque (Area 1), the
ICE torque should be decreased bringing it near the optimal torque point. It means that EM
should be run as a motor to make up for the remaining torque, provided there is enough
battery charge.
At the same current speed, if the required torque is below the optimal torque (Area 2), the ICE
torque should be increased bringing it near to the optimal torque point. This is possible only if
SOC is not high. We can run the EM as a generator, while running the ICE at its optimum.
In order to modeling, the following specification of engine and Motor/Inverter will be
considered. Figs 9 and 10 show that the fuel converter efficiency operation and as well
Motor/Inverter Efficiency.
Optimal Tourque Area
(Fuel Efficiency)
0.3
0.35
0.3
Fuel Efficiency
0.25
0.25
0.2
0.15
0.1
90 0.2
80
70
60
50 0.15
40 6000
Tourqe(N/m) 30
5000
4000
20 3000
10 2000
0
1000 Speed(RPM)
0
Fig. 9. Fuel Converter Operation Honda Insight 1.01 VTEC-E SI from ANL Test Data .
Motor Efficiency 0.95

Area
Generator Efficiency 0.9

Area
1 0.85
otor/Inverter Efficiency
0.9
0.8
0.8
0.75
0.7
0.7
0.6
M
0.65
0.5
60
0.6
40 9000
8000
20 7000
6000 0.55
0 5000
4000
-20 3000
2000
1000
-40 0
Motor Torque(Nm) Motor Speed(RPM)
Fig. 10. Motor/Inverter Efficiency and Condition Torque Capability (Preliminary Model of
Honda 10kw).
www.intechopen.com
3.4.4 On-line adaptive strategy

The general control strategy for a parallel HEV can be summarized as follows (Shi et al,.2006):
i. When the speed of the vehicle is small, ICE stops and electric motor gives the
driving power required which avoids higher fuel consumption and reduce
emission (It is assumed that SOC is sufficient).
ii. When the speed of the vehicle is high enough, electric motor stops, ICE starts and
gives the driving power required. Currently, ICE works along optimum curve
depending on the cost function.
iii. If the power required is larger than what ICE can give, ICE and electric motor work
together and electric motor takes additional required power from the battery (It is
assumed that SOC is sufficient).
iv. If SOC of the battery drops under the safe level, ICE supplies both the energy
required for travelling and extra power to charge the battery through electric motor
(electric motor is at generator mode).
v. In brake state, energy floats from vehicle body to drivetrain. Electric motor works
as a generator and transforms braking energy to electricity to charge the battery.
4. Intelligent System Methods in Energy Management

Intelligent energy management methods can observe and learn driver behavior,
environmental and vehicle conditions, and intelligently control the operation of the hybrid
vehicle. This section describes intelligent system approaches with applications to design
optimization, modeling, and control of complex systems and processes.
4.1 Introduction of Complex and Uncertain System

A Complex System is (Simon. H,1973) “A system that can be analyzed into many
components having relatively many relations among them, so that the behaviour of each
component depends on the behaviour of others”.
In the real world, we can find many problems and systems that are too complex or uncertain
to be represented in complete and accurate mathematical models. And yet, we still have the
need to design, optimize, or control the behaviour of such systems. Complex system can be
solved by artificial intelligent systems.
Advances in intelligent systems have brought new opportunities and challenges for
researchers to deal with complex and uncertain problems and systems, which could not be
solved by traditional methods. Methods developed for mathematically well-defined
problems with precise models may lack in autonomy and thus cannot give adequate
solutions under uncertain environments (Shin &Xu, 2009). Intelligent systems are defined
with high degree of autonomy, reasoning with uncertainty, higher performance, high level
of abstraction, data fusion, learning and adaptation (Shoureshi & Wormley,1990).
4.2 Soft Computing Techniques

Various soft computing based techniques have emerged as useful tools for solving
engineering problems that were not possible or convenient to handle by traditional
methods. The soft computing techniques give computationally efficient modelling, analysis,
and decision making. The techniques that belong to the soft computing include artificial
neural networks (ANNs), Fuzzy sets and systems, and evolutionary computation.
www.intechopen.com
4.2.1 Artificial neural networks (ANNs)

ANNs are collections of small individually interconnected processing units. Information is
passed between these units along interconnections. An incoming connection has two values
associated with it, an input value and a weight. The output of the unit is a function of the
summed value. ANNs while implemented on computers are not programmed to perform
specific tasks. Instead, they are trained with respect to data sets until they learn patterns
used as inputs. Once they are trained, new patterns may be presented to them for prediction
or classification. ANNs can automatically learn to recognize patterns in data from real
systems or from physical models, computer programs, or other sources. They can handle
many inputs and produce answers that are in a form suitable for designers.
4.2.2 Genetic Algorithms (GA)

GA is based on the way living organisms adapt to life by evolution and inheritance. GA
imitates the process of evolution of population by selecting fit individuals for reproduction.
Thus, GA is an optimum search technique based on the concepts of natural selection and
survival of the fittest. It works with a fixed-size population of possible solutions of a
problem, called individuals, which are evolving in time. A genetic algorithm utilizes three
principal genetic operators: selection, crossover, and mutation.
4.2.3 Fuzzy Logic (FL)

FL is used in control engineering. It is based on reasoning which employs linguistic rules in
the form of IF-THEN statements. FL provides a simplification of a control methodology
description. This allows the human language to be used to describe the problem and its
solutions. In many control applications, the model of the system is unknown or the input
parameters are variable and unstable. In such cases, fuzzy controllers can be applied. These
are more robust and cheaper than conventional PID controllers. It is also easier to
understand and modify fuzzy controller rules, which not only use human operator’s
strategy but, are expressed in natural linguistic terms.
4.2.4 Hybrid system (HS)
Techniques Advantage Limitation

Expert Systems also called -Cost reduction in achieving a -The lack of expertise
Knowledge-Based Systems (KBS) complex task
Artificial Neural Networks -Most of the problems are now -Is only a special mathematical
(ANNs) able to be solved technique
-Representing I/O relationships
for nonlinear systems.
Fuzzy logic (FL) -Applied successfully in large -Input/output controls of
number of uncertain applications process are complicated
Genetic Algorithms (GAs) -Optimisation -Successful for some
applications
Hybrid System -Combination technique is -No
(ANNs & FLS), (FLS & ANNs), capable to solve all problems
(GAs & FLS) and (GAs & ANNs) of engineering discipline
Table 4. Intelligent system methods.
www.intechopen.com
HS combines multiple soft computing methods. For example, neuro-fuzzy controllers use neural
networks and fuzzy logic, whereas in a different hybrid system a neural network may be used to
derive some parameters and a genetic algorithm may be used subsequently to find an optimum
solution to a problem. Table 4 presents a comparison of features of soft computing methods.
5. Proposed Intelligent Energy Management System

This work employs the analysis and simulation approach to develop an Intelligent Energy
Management System (IEMS) for a HEV. The overview of IEMS is shown in Fig12. IEMS
calculates the energy distribution and power flows in the powertrain of the vehicle and
related losses. It indicates the ways to minimises the vehicles’ fuel consumption under
various driving conditions. IEMS learns when it is run, and makes proper adjustments to the
way it operates to ensure that fuel consumption optimisation is achieved.
The developed model includes the following components:
5.1 Look-Ahead Environment Model Unit (LEM):

This unit employs an imaging sensor and a vision algorithm to calculate the slope angle of
the road ahead within the distance of 300 meters away from the vehicle, and forward this
information to IEMS.
5.2 Current Environment Model Unit (CEM):

This unit employs the following data from environment situation.
i. Current Road Slope Module (CRSM): This module specifies the actual slope angle of
the road at the current location of the vehicle.
ii. Road Friction Module (RFM): This module gives road friction coefficient, gravity
acceleration, and motion angle.
iii. Wind Drag Module (WDM): This module provides the following wind parameters:
wind speed, wind direction, and drag coefficient.
5.3 Friction Management Unit (FMU):

This module obtains CEM data and also the following data to calculate and send them to IEMS.
i. Combustion Module (CM): This module employs the combustion process from the
vehicle as described in Equation (6-8), and calculates and returns the amount of
combustion energy needed.
ii. Accessory Module (AM): This module represents the accessories embedded within
the vehicle such as electrical devices and air conditioning.
iii. Vehicle Efficiencies Module (VEM): This module defines the values of the otto cycle,
real fuel air engine, mechanical and heat loss efficiencies.
5.4 Battery State of Charge (SOC):

This module provides the amount of current, temperature and voltage of the battery
continuously. Figure 11 displays only that the useable area of charge on the hybrid battery,
displayed "empty" is about 40% and displayed "full" is about 85%.
www.intechopen.com
Fig. 11. The useable area of charge and discharge on the hybrid battery.
5.5 Control Strategy

In this work, the on-line adaptive strategy which was discussed in section 3.4.4, has been
considered.
5.6 IEMS Algorithm

The overview of the simulation algorithm for IEMS is displayed in Fig. 13. The simulation
starts with initialising several variables including normal power and primal kinetic energy
for a moving vehicle. The data includes arrays of 7200 elements (steps). One iteration occurs
in each step representing the time interval of 0.05 sec. Then the slope prediction data is
retrieved from LEM. If the predicted slope angle is different from the current slope angle,
STI block increases or decreases the power. Next, the vehicle/environment/friction data is
retrieved from FMU. If the current total friction energy is different from the energy
associated with the slope prediction, FTI block is triggered calculating the amount of power
for all frictions. Otherwise, fuzzy logic controller (FLC) block is entered. FLC controls and
optimises the fuel consumption with respect to the vehicle/efficiencies, speed, acceleration,
and gear data. Also the FLC intelligently consider with drive strategy (see section 5.7) and
control strategy. If the comparison is satisfied then these data will be forwarded to the next
block where they overwrite the results of the previous iteration. Otherwise, the power of
engine and inverter operation is corrected by decreasing or increasing. Once either of speed
or acceleration is found to be greater than the desired limit, and then FLC will control the
engine power and inverter operation by its algorithm. When either of speed or acceleration
becomes smaller than the desired control strategy limit, the engine power is increased with
regard to control strategy. In the assignment block, the old data is overwritten with the new
data. The Inverter algorithm, shown in Fig. 14, synchronises the battery with EM, Gen. With
regard to the battery SOC and IEMS Interpreter Load (IIL), inverter starts charging or
discharging the battery in each time. It then informs the IEMS about its result via SOC. If
SOC is high, and at times of high load, the generator can be switched off and EM can
provide mechanical power via the battery by the inverter’s instruction. The parameters of
the FLC controller are optimized by genetic algorithm optimizer (GAO).
www.intechopen.com
Fig. 12. Overview of the IEMS model.
| Slopef  SlopeC | 0
| f t  f t 1 | 0
t  t 1
Pt 1  Pt  PICE  PE
a t 1  a t
Vt 1  Vt
Fig. 13. Overview of the IEMS algorithm.
www.intechopen.com
Fig. 14. Overview of the inverter algorithm.
5.7 Engine and EM/Gen Specification and Drive Strategy

In this work, we have considered a vehicle with the engine and specification as given in Table 5.
Parameters Min Max Average

Engine size (litre) ---- 1.1 ---------
Engine RMP (Rev/min) 3000 4000 3500
Engine power (kW) 9.8 10.5 10
Engine Torque (N/m) 25 29 27
EM/Gen RPM 3000 4000 3500
(Rev/min)
EM/Gen Torque(N/m) 8 12 10
Table 5. Engine and EM/Gen specification.
We have also formulated a set of parameters called “Drive Strategy” as shown in Table 6.

Engine size (litre) ---- 1.2 ---------
Speed (m/s) 16.38 16.94 16.66
Acceleration (m/s2) -0.98 0.98 0.5
Travel Time ------ 7200 0.05(s)
Travel Distance (m) 6000
Table 6. Drive strategy parameters.
www.intechopen.com
6. Simulation and optimization of hybrid vehicle

6.1 Simulation
6.1.1 Simulation 1
Khayyam et al (Khayyam et al, 2009b) demonstrated a Air Condition system simulation that,
the vehicle was tested under sunny condition first for 1200 step for the vehicle speed around
20 m/s. Next, the fan and the air conditioning are turned on. The parameters given in Tables
7-9 were employed to achieve the comfort temperature in the cabin room for 6000 step. The
air condition energy consumption shown in Fig 15(d).
Parameters Min Max Normal

C.O.P 1.45 1.71 1.38
CAP (KW) 3.8 8.15 7.9
RMP ( rev / min ) 3000 4000 3500
Evaporator (KW) 5.51 13.93 10.90
Temperature (C) 0 10
Gas R-134 Sub Cool superheat
Pressure (kPa) 310 2415
Gas R-134 Charge Discharge
Table 7. Compressor specifications.

Volts 12.5 12.6 12.5
Amp 20 25 24
RMP ( rev / min ) 1000 1800 1200
Engine power (W) 250 315 300
Table 8. Blower specifications.

Temperature (C) 20 25.6 21.5
Humidity (%) 40 60 50
Air speed (m/s) 1 5 2.5
Table 9. Comfort cabin room specifications.
As discussed in section 3.2, some data has been created by PMH technique. The data created
is associated with a slopped-windy-sunny condition. HEV was tested on this data, where
the hybrid electric components were included. The road was set to be slopped with various
slope angles within the range – 1% ≤ atan(Ø) ≤ +0.6%. Moreover, the environmental wind
was assumed to be non-zero. The wind angle of attack, θ2, was varied within the range 0 to
360°. Considering the wind velocity, however, different conditions were implemented: V2=0
to 6. The Current Environment Model (CEM) component monitored the current slope. The
following parameters were also considered: road-friction, combustion, and air conditioning
accessory (Table1). Fig. 15 illustrates the slope angle, wind-speed, wind direction as well as
A/C energy consumption data used in the simulation.
www.intechopen.com
0.6
0.4
0.2
Slope(%)
-0.2
-0.4
-0.6
-0.8
-1
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000
Step
(a)
7
5
Wind Speed
4
(m/s)
0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 70007200
Step
(b)
350
300
Wind Direction
250
(degree)
200
150
100
50
0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 70007200
Step
(c)
4000
3000
Consumption
A/C Energy
(Watt)
2000
1000
0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000
Step
(d)
Fig. 15. (a) Slope (Road) angle data, (b) wind-speed data ,(c) wind-direction and (d) A/C
energy consumption.
6.1.2 Simulation 2
IEMS-HEV was tested on a set of data associated with a slopped prediction (look-ahead
within a 300 meter distance)-windy-warmed employing the hybrid electric components. The
management of the battery, EM, and Gen is conducted by the inverter algorithm. This
enables ICE and EM to output power simultaneously when the load is greater than 10 kW or
a slope of greater than 0.1% is climbed by the vehicle. The following parameters were also
www.intechopen.com
considered: road-friction, combustion, and air conditioning accessory. The predicted slope
angle data is similar to the actual slope angle data.
6.2 Discussions
6.2.1 Simulation 1 Results

The power and fuel consumption results for the first simulation are shown in Fig. 16.
Initially. 7800 W of energy is given to the vehicle so that the initial speed of 16.6 m/s is
achieved. The energy consumption remained constant at 7800 W where the condition was
flat-windless (e.g. steps 0-600). Depending on the condition of the road slope angle, the
wind speed, angle of attack, and accessory energy consumption, the power consumption
varied as shown in Fig. 16. The HEV was informed about the current slope by CEM.
Fig. 16 shows that the air conditioning system and slope friction have a significant impact on
the fuel consumption. The reason is that it requires more fuel in a transit time. HEV can
measure how much energy is needed in each step, and works out a desired fuel rate for the
engine so that the power brake would not be needed. Using Equation (10), the average fuel
consumption for Simulation 1 was found to be around 6.65 liter/100 km.
4 -3
x 10 x 10
1.5 1.5
1.4
P o w e r C o n s u m p tio n
F u e l C o n s u m p tio n
1.3
1.2
1.1
(W a tt)
(lite r)
1 1
0.9 Power Consumption
Fuel Consumption
0.8
0.7
0.6
0.5 0.5
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000
Step
Fig. 16. Power and fuel consumption results for Simulation 1.
6.2.2 Simulation 2 Results

The power and fuel consumption results for the second simulation are shown in Fig. 17.
Similarly, 7800 W of energy is initially given to the vehicle so that the initial speed of 16.6
m/s is achieved. Also, the Look-Ahead informed IEMS about any slope ahead. IEMS
calculates and FLC investigates, and if the load is found to be greater than 10 kW or the
slope greater than 0.1%, the propulsion balance requests ILL to switch on EM through
inverter. The outcome of this simulation shows that the vehicle speed and acceleration are
smoother than those of Simulation 1. Using Equation (10), the average fuel consumption for
Simulation 2 was found to be around 6.11 liter/100 km. Fig. 18 shows the SOC of battery
during the travel. It goes up to 85% and then comes down to the same level when used.
www.intechopen.com
12000
10000
Power Consumption 8000 Required Power

EM/Gen Power
6000
(W att)
ICE Power
4000
2000
-2000
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000
Step
Fig. 17. Power consumption results for Simulation 2.
1
0.9
0.8
0.7
SOC
0.6
0.5
0.4
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000
Step
Fig. 18. SOC of battery.
7. Conclusions
This chapter presented a description of intelligent energy management systems for hybrid
electric vehicles. In addition, an intelligent energy management model for a parallel hybrid
electric vehicle was described. The model takes into account the role of combined
wind/drag, slope, rolling, and accessories loads to minimize the fuel consumption under
various driving conditions. Two simulation studies were conducted. They show that the
vehicle speed and acceleration were smoother when the hybrid section was included. The
average fuel consumption for Simulation 1 and 2 were found to be around 6.65 and 6.11
liter/100 km, respectively.
8. References
A. Bandivadekar and J. Heywood,(2007) "Factor of two : halving the fuel consumption of
new U.S. automobiles by 2035," Report from Laboratoy for energy and the
environment (MIT), vol. LFEE,.
A. Piccolo, L. Ippolito, V. Galdi, and A. Vaccaro,(2001) "Optimization of energy flow
management in hybird electric vehicles via genetic alorithms," Proc. IEEE/ASME
Int. Conf. Italy, pp. 434-439,.
A. Sciarretta, M. Back, and L. Guzzella,(2004)"Optimal Control of Parallel Hybrid Electric
Vehicles," IEEE Trans. on Control Systems Technology, vol. 12,.
B. M. Baumann, G. Washington, B. C. Glenn, and G. Rizzoni,(2005) "Mechatronic design and
control of hybrid electric vehicles," IEEE/ASME Transactions on Mechatronics, vol.
5, pp. 58-72, 2000.
C. Musardo, G. Rizzoni, Y. Guezennec, B Staccia.(2005)“A-ECMS: An Adaptive Algorithm
for Hybrid Electric Vehicle Energy Management”; European Journal of Control,
Fundamental Issues in Control, Special Issue, Volume 11, Number 4-5.
www.intechopen.com
C. C. Lin, H. Peng, and J. W. Grizzle,(2003) "Power management strategy for a parallel

hybrid electric truck," IEEE Trans. on Control Systems Technology, vol. 11, pp. 839-
849,.
C. C. Linl, H. Pengl, and J. W. Grizzle, (2004)"A Stochastic Control Strategy for Hybrid
Electric Vehicles," Proceeding of the 2004 American Control Conference Boston,
Massachusetts, vol. 4710-4715,.
E. D. Tate and S. P. Boyd,(1998) "Finding ultimate limits of performance for hybrid electric
vehicles," SAE, vol. 00FTT-50,.
E. Hellstrom, M. Ivarsson, J. Aslund, and L. Nielsen,(2009) "Look-ahead control for heavy
trucks to minimize trip time and fuel consumption," control Engineering Practice,
vol. 17, pp. 245-254,.
EPA, (2004)"Assessing the emission and fuel consumption impacts of intelligent
transportation system(ITS)," EPA, vol. 231, pp. 98-007,.
G. J. v. Wylen and R. E. Sonntag,(1990)"fundamentals of classical thermodynamics " vol.
Third edition SI version.
G. Shi, Y. Jing, A. Xu, and J. Ma, (2006)"Study and Simulation of Based-fuzzy-logic Parallel
Hybrid Electric Vehicles Control Strategy," Proceedings of the Sixth International
Conference on Intelligent Systems Design and Applications pp. 280-284,.
H. Khayyam, A. Z. Kouzani, and E. J. Hu,(2008) "An intelligent energy management model
for a parallel hybrid vehicle under combined loads," in IEEE International
conference on vehicular electronics and safety Columbus, OH, USA 2008, pp. 145-
150.
H. Khayyam, A. Z. Kouzani, and E. J. Hu,(2009b) "Reducing energy consumption of vehicle
air conditioning system by an energy management system," IEEE The 4th
Internationnal Green Energy Conferance Beijing, China, vol. 1
H. Khayyam, A. Z. Kouzani, H. Abdi, and S. Nahavandi,(2009a) "Modeling of highway heits
for vehicle modeling and simulation " Proc. ASME/IEEE Int. Conf. MESA09 USA.
J. B. Heywood, (1998)"Internal combustion engine fundamentals," McGraw-Hill, New York,.
J. S. Won, (2003)."Intelligent energy management agent for a parallel hybrid vehicle," PhD
Thesis 2003.
J. T.B.A. Kessels, (2007) "Energy Management for Automotive Power Nets," PhD thesis 2007.
L. L. Yaodong Wang, (2007)"An analytic study of applying Miller cycle to reduce Nox
emission from petrol engine " Applied Thermal Engineering, vol. 27, pp. 1779-1789,.
Lewis,Frank L. & Syrmos, Vassilis L. (1995) “Optimal Control” Hohn Wiley & Sons Inc.
Newyork.
M. Ehsani, Y. Gao, S. E. Gay, and A. Emadi, (2005) “Modern electric, hybrid electric, and fuel
cell vehicles”, New York CRC PRESS.
M. Farrokhi and M. Mohebbi, (2005) "Optimal Fuzzy Control of Parallel Hybrid Electric
Vehicles," ICCAS2005 June 2-5.
M. Mohebbi and M. Farrokhi,(2007) "Adaptive neuro control of parallel hybrid electric
vehicles," Int. J. Electric and Hybrid Vehicles, vol. 1,.
M. Mohebbi, M. Charkhgard, and M. Farrokhi,(2005)"Optimal neuro-fuzzy control of
parallel hybrid electric vehicles," Vehicle Power and Propulsion, 2005 IEEE
Conference, pp. 26-30.
M. Plint and A. Martyr,(1997) "Engine testing theory and practice," Butter worth-Heinemann
British Library .
www.intechopen.com
M. Salman, F. Change, and J. S. Chen,(2005) "Predictive energy management strategies for

hybrid vehicles," IEEE Conf. Decision Control pp. 21-25,.
M. Salman, N. J. Schouten, and N. A. Kheir, (2000) "Control strategies for parallel hybrid
vehicles", American Control Conference, vol. 1, pp. 524-528.
Manins ,P.C.(2000)."Environmental impact of the new Australian Hybrid Cars " Proceedings
of the 15th International Clean Air and Environment Conference vol. 26-30, pp.
117-122,.
N. J. Schouten, M. A. Salman, and N. A. Kheir,(2002)"Fuzzy logic control for parallel hybrid
vehicles," IEEE Transactions on Control Systems Technology, vol. 10, pp. 460-468, .
O. Barbarisi, E. R. Westervelt, F. Vasca, and G. Rizzoni,(2005) "Power management
decoupling control for a hybrid electric vehicle," IEEE Conf on Decision and
Control, Spain, pp. 2012-2017.
P. C. Cacciabue, (2007)"Modelling driver behaviour in automotive environments “: Springer
2007.
P. Pisu, G. Rizzoni, (2007) "A Comparative Study of Supervisory Control Strategies For
Hybrid Electric Vehicles" IEEE Transactions on Control Systems Technology,
Special Issue on Control Applications in Automotive Engineering,
Q. R. Riley, (1994)"Alternative cars in the 21 century a new personal transportaion
paradigm," SAE International, vol. 1994,.
R. Langari and J.-s. W. Won, (2005)"Intelligent Energy Management Agent for a Parallel
Hybrid Vehicle-Part 1 : Systsem Architecture and Design of the Driving Situation
Identification Process," IEEE Trasactions on Vehicular Technology vol. 54, pp. 925-
934,.
R. Shoureshi and D. Wormley,(1990) "Intelligent control systems," Final report of
NSF/EPRI,.
R. Zhang and Y. Chen, (2001) "Control of hybrid dynamical systems for electric
vehicles,"American Control Conference, vol. 4, pp. 2884-2889,
S. Barsali, C. Miulli, and A. Possenti, (2004)"A control strategy to minimize fuel
consumption of series hybrid electric vehicles," IEEE Transaction on Energy
Conversion, vol. 19, pp. 187-195,
S. Delprat, J. Lauber, T. M. Guerra, and J. Rimaux,(2004) "Control of a Parallel Hybrid
Powertrain: Optimal Control," IEEE Transaction on vehicular technology, vol. 53,
pp. 872-881,.
S. E. Lyshevski,(1999)"Diesel-Electric Drivetrains for Hybrid-Electric Vehicles: New
Challenging Problems in Multivariable Analysis and Control," Proceedings of the
1999 IEEE International Conference on Control Applications vol. 1, pp. 840-845,.
S. M. T. Bathaee, A. H. Gastaj, S. R. Emami, and M. Mohammadian,(2005) "A fuzzy-based
supervisory robust control for parallel hybrid electric vehicles," Vehicle Power and
Propulsion, 2005 IEEE Conference, pp. 694-700,.
Simon, Herbert A., (1973) "The Organization of Complex Systems," in Hierarchy Theory, ed.
by Howard H. Pattee, New York: George Braziller,
SOE,(2006) "Independent report to the Australian Government Minister for the Environment
and Heritage " <http://www.environment.gov.au/soe/2006/publications/report/
index.html>, 2006.
W. W. Pullkrabek,(1997)"Engineering fundamentals of the internal combustion engine "
Prentice Hall.
www.intechopen.com
X. Wei, L. Guzzella*, V. I. Utkin and G. Rizzoni, (2007) “Model-based Fuel Optimal Control
of Hybrid Electric Vehicle Using Variable Structure Control Systems”, ASME
Transactions. Journal of Dynamic Systems, Measurement, and Control - Volume
129, Issue 1, pp. 13-19.
Y. C. Shin and C. Xu,(2009)” Intelligent Systems modeling, optimization, and control” Boca
Raton : CRC Press,.
Y. Lin, P. Tang, W. J. Zhang, and Q. Yu,(2005)"Artificial neural network modelling of driver
handling behaviour in a driver-vehicle-environment system," International Journal
of Vehicle Design, vol. 37, pp. 24-45,.
Y. Wang, L. Lin, and A. roskily(2007). "An analytic study of applying Miller cycle to reduce
Nox emission from petrol engine " Applied Thermal Engineering, vol. 27, pp. 1779-
1789
Y. Zhu, Y. Chen, and Q. Chen,(2002) "Analysis and Design of an Optimal Energy
Management and Control System for Hybrid Electric Vehicles," Proc. of the 19th
Electric Vehicles Symposium, Busan, Korea,.
www.intechopen.com
www.intechopen.com
Energy Management
Edited by Francisco Macia Perez
ISBN 978-953-307-065-0
Hard cover, 246 pages
Publisher InTech
Published online 01, March, 2010
Published in print edition March, 2010
Forecasts point to a huge increase in energy demand over the next 25 years, with a direct and immediate
impact on the exhaustion of fossil fuels, the increase in pollution levels and the global warming that will have
significant consequences for all sectors of society. Irrespective of the likelihood of these predictions or what
researchers in different scientific disciplines may believe or publicly say about how critical the energy situation
may be on a world level, it is without doubt one of the great debates that has stirred up public interest in
modern times. We should probably already be thinking about the design of a worldwide strategic plan for
energy management across the planet. It would include measures to raise awareness, educate the different
actors involved, develop policies, provide resources, prioritise actions and establish contingency plans. This
process is complex and depends on political, social, economic and technological factors that are hard to take
into account simultaneously. Then, before such a plan is formulated, studies such as those described in this
book can serve to illustrate what Information and Communication Technologies have to offer in this sphere
and, with luck, to create a reference to encourage investigators in the pursuit of new and better solutions.
How to reference
In order to correctly reference this scholarly work, feel free to copy and paste the following:
Hamid Khayyam, Abbas Kouzani, Saeid Nahavandi, Vincenzo Marano and Giorgio Rizzoni (2010). Intelligent
Energy Management in Hybrid Electric Vehicles, Energy Management, Francisco Macia Perez (Ed.), ISBN:
978-953-307-065-0, InTech, Available from: http://www.intechopen.com/books/energy-
management/intelligent-energy-management-in-hybrid-electric-vehicles
InTech Europe InTech China

University Campus STeP Ri Unit 405, Office Block, Hotel Equatorial Shanghai
Slavka Krautzeka 83/A No.65, Yan An Road (West), Shanghai, 200040, China
51000 Rijeka, Croatia
Phone: +385 (51) 770 447 Phone: +86-21-62489820
Fax: +385 (51) 686 166 Fax: +86-21-62489821
www.intechopen.com
© 2010 The Author(s). Licensee IntechOpen. This chapter is distributed
under the terms of the Creative Commons Attribution-NonCommercial-
ShareAlike-3.0 License, which permits use, distribution and reproduction for
non-commercial purposes, provided the original is properly cited and
derivative works building on this content are distributed under the same
license.

Intelligent Energy Management in Hybrid Electric Vehicles

Uploaded by

Copyright:

Available Formats

Intelligent Energy Management in Hybrid Electric Vehicles

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Intelligent Energy Management in Hybrid Electric Vehicles

Uploaded by

Copyright:

Available Formats

Intelligent Energy Management in Hybrid Electric Vehicles 147

Intelligent Energy Management

Fig. 1. Three HEVs structures.

large number of mechanical, electronic, and electromechanical elements (Zhu et al.,2002).

2. Literature Review and Background

2.1 Rule-Based Control

2.2 Optimal Control

A. Global Optimization (off line)

A.1 Neural Networks

A.2 Classical Optimal Control

A.3 Linear Programming

A.4 Dynamic and Stochastic Programming

A.5 Genetic Algorithm

B. Real Time Optimization (on line)

B.1 Model Predictive Control

B.2 Decoupling Control

3. Factors Involved in Energy Management of Hybrid Electric Vehicles

3.1 Main factors involved in energy management system of HEVs

Fig. 2. Overview of the energy management model for HEVs.

3.2 Model of Environment

Fig. 3. Highway creation algorithm using the developed PHM.

Fig. 5. Wind creation algorithm using the developed PHM.

3.3 Model of Driver

Acceleration Calm Normal Aggressive

3.4 Model of Vehicle (Quasi-Static)

Fig. 6. P-HEV topology (Kessels.J,2007).

3.4.1 Conventional Vehicle Specification

Q combaustion  (Q fuel  Q air  Q exhaust ) (3)

( Q combustion )   otto   fuel  air

where: Pnet = Power output of engine

otto =Otto cycle efficiency=

 = Real fuel air engine efficiency =0.75(Yaodong Wang,2007)

 =Mechanical efficiency=0.9 (Plint ,1997)

 = Heat loss efficiency=0.8 (Pullkrabek,1997)

Q combustion  m fuel  qcombustion (5)

C 8 H 14.96  k  11 .74  (O2  3.76 ) N 2 

Type Symbol Value

Sensible formation E-F Thermodynamic tables

Mass flow rate of fuel V-O . kg

Temperature of fuel V-S T fuel 27 °C

Proad friction  Pdrag  Pslope  Paccessory  Pdriving

The total fuel consumption in this process is:

Pnet  [ F friction  F drag  F slope  F ] Vt  [1 / 2  m  (V t 2 1  V t )]

where t is the total numbers of steps involved in the simulation.

Also, the distance traversed by vehicle in t is:

3.4.2 Parallel Hybrid Electric Vehicle Specification

Fig. 7. Battery Model

Feature Symbo Type

3.4.3 Control strategy and optimal torque

Fig. 8. Optimal torque map

Motor Efficiency 0.95

Generator Efficiency 0.9

3.4.4 On-line adaptive strategy

4. Intelligent System Methods in Energy Management

4.1 Introduction of Complex and Uncertain System

4.2 Soft Computing Techniques

4.2.1 Artificial neural networks (ANNs)

4.2.2 Genetic Algorithms (GA)

4.2.3 Fuzzy Logic (FL)