MODELING AND ANALYSIS

Multiobjective optimization for hydro-­photovoltaic hybrid

power system considering both energy generation and
energy consumption
Fang-Fang Li1,2,3, Jun Qiu2 & Jia-Hua Wei2,3
1College of Water Resources & Civil Engineering, China Agricultural University, 100083, Beijing, China
2State Key Laboratory of Hydroscience & Engineering, Department of Hydraulic Engineering, Tsinghua University, 100084, Beijing, China
3State Key Laboratory of Plateau Ecology and Agriculture, Qinghai University, 810016, Xining, China

Keywords Abstract
Energy consumption, energy generation,
hydro-photovoltaic hybrid system, Hydropower can be an ideal compensation for fluctuant photovoltaic (PV) power
multiobjective optimization, NSGA-II due to its flexibility. In this study, a multiobjective optimization model con-
sidering energy generation and consumption simultaneously for a hydro-­ PV
Correspondence hybrid power system is proposed. With the daily mean radiation intensity and
Jia-Hua Wei, State Key Laboratory of temperature, the PV power output is calculated. Taking reservoir release as the
Hydroscience & Engineering, Department of
decision variable, the total energy generation of the hydro-­PV system is maxi-
Hydraulic Engineering, Tsinghua University,
100084 Beijing, China. E-mail: weijiahua@
mized. Meanwhile, the gap between the energy generation and the energy con-
mail.tsinghua.edu.cn sumption is minimized to reduce the abandoned PV power or hydropower.
The proposed multiobjective model is optimized by Non-­ dominated Sorting
Funding Information Genetic Algorithms-­II (NSGA-­II). The Longyangxia Project, the largest hydro-­
National Key R&D Program of China, Grant/ photovoltaic hybrid power system in the world is taken as the study case to
Award Number : 2017YFC0403600 and estimate the optimal operational strategies for different objectives in wet year,
2017YFC0403602; Open Research Fund
normal year, and dry year, respectively. The optimal operation process of the
Program of State key Laboratory of
Hydroscience and Engineering, Grant/Award
reservoir is presented, which is instructive for the operation in the future.
Number: sklhse-2016-B-03; the Open Project
of State Key Laboratory of Plateau Ecology
and Agriculture, Grant/Award Number:
2016-KF-03

Received: 5 September 2017; Revised: 14

May 2018; Accepted: 15 May 2018

doi: 10.1002/ese3.202

Introduction the cumulative installed Photovoltaic (PV) capacity had

exceeded 77 GW, ranking the first in the world. Its an-
Solar energy is the most abundant and widely distributed nual power generation was 66.2 billion kWh, taking up
renewable energy resource on the earth, having a prospect 1% of the total amount of national electric energy pro-
of wide development and utilization. With the technology duction [2].
progress and the cost reduction, solar energy has attracted PV output is affected by solar radiation and temperature
more and more attention all over the world. with obvious diurnal and seasonal cycles. In addition,
The annual solar radiation on 2/3 of China’s territory affected by weather variations, PV output has strong in-
is over 5000 MJ/m2, and the annual solar energy absorbed termittency. With the increasing proportion of the PV
by the surface is equivalent to 170 billion tons of standard power in the power grid, the impact of PV power gen-
coal energy [1]. The latest data from National Energy eration on the planning, security, scheduling, and control
Administration of China shows that by the end of 2016, of the power system becomes more and more notable

© 2018 The Authors. Energy Science & Engineering published by the Society of Chemical Industry and John Wiley & Sons Ltd. 1
This is an open access article under the terms of the Creative Commons Attribution License, which permits use,
distribution and reproduction in any medium, provided the original work is properly cited.
Multi-objective optimization of hydro-PV F.-­F. Li et al.

[3]. The power supply system needs to have enough re- [13]. Schmidt et al. concluded that the existing reservoirs
serve capacity and adjustment measures to mitigate such of hydropower in Brazil are sufficient to balance varia-
impact. Therefore, it is necessary to seek effective means tions in renewable electricity supply at an optimal mix
to reduce the impact of PV power fluctuations on the of around 37% of PV, 9% of wind, and 50% of hydro-
operation of power grid. power generation [14]. Kougias et al. optimized the in-
As another important energy sources, hydropower has stallation characteristics of the solar PV system to improve
fast regulation ability, and often plays a role of peak the total energy production of a hybrid solar PV and
regulation, frequency modulation, and emergency standby small hydropower system [15].
in the power system. Such characteristics can be used to In this study, we attempt to take the energy generation
compensate the PV power output fluctuation to provide and consumption of the hybrid hydro-­ PV system into
high-­quality electric energy for power grid. account simultaneously, and a multiobjective optimization
The first 10 MW-­level hydro-­PV complementary power model maximizing energy generation and minimizing the
station in the world was built in 2009 in Yushu of Qinghai gap between the energy production and consumption
province in China [4]. In 2015, the Longyangxia project, energy for a hydro-­PV hybrid power system is proposed.
which is the world’s largest hydro-­ PV power plant was Taking reservoir release as the decision variable, and con-
completed and put into operation in Qinghai project with sidering the multiannual means of daily radiation and
the total PV installed capacity of 850 MW [5]. temperature, the total amount of annual energy generation
There have been some researches and attempts on of the hydro-­ PV system is maximized; meanwhile, the
hydro-­PV hybrid power systems, including its plan, design, gap between energy generation and consumption is mini-
construction, and control. Wu et al. analyzed the opera- mized. The proposed multiobjective model is optimized
tion characteristics of small hydro/photovoltaic power by NSGA-­II. The optimal operational strategies in different
system, and proposed a novel control strategy [6]. Meshram hydrologic years for the Longyangxia Project are estimated.
et al. simulated and modeled standalone DC linked hydro/ Not only the trade-­off relationships between the two ob-
PV/battery hybrid energy system and power management jectives are presented, but also the corresponding opera-
strategy for identifying the active power sharing [7]. tional strategies are analyzed, which is instructive for the
Kenfack et al. simulated and sized a microhydro-­ PV-­ operation in the future.
hybrid system for rural electrification [8]. Bekele et al.
and Nfah et al. studied feasibility of small-­scale Hydro/
PV/Wind based hybrid electric supply system for off-­grid Methodology
rural electrification in Ethiopia [9] and pico-­ hydro/PV
hybrid power system for remote villages in Cameroon PV power output model
[10], respectively. Besides these researches concentrating Using the dual-­ axis tracking device and the Maximum
on the control techniques of small-­scale, off-­grid hydro/ Power Point Tracking (MPPT) technique, the output of
PV hybrid systems, studies on optimizing or evaluating the PV array is only affected by the solar radiation in-
the operation of the hybrid Renewable Energy Systems tensity and environmental temperature, where temperature
(RES) were also carried on. Beluco et al. evaluated of is a parameter of the efficiency of the solar PV modules
the use of water reservoirs and battery banks as alterna- in equation (13). The output voltage at the maximum
tives for energy storage within a hybrid hydro-­PV system, power point can be approximately considered to be con-
and proposed the concept of a theoretical performance stant. Hence, the basic output model of PV power gen-
limit for the plants and propose a method for the de- eration system obtained by National Renewable Energy
termination of this performance limit using computer Laboratory (NREL) [16] is:
simulations with idealized energy availability functions
[11]. Patsialis et al. demonstrated that hydropower can ( )
provide additional flexibility to the local system and �
RT
PPV = P PV
[1 + 𝛼P (TC − TSTC )], (1)
through reservoir operation to allow the connection of RSTC
additional solar photovoltaic capacities [12]. François et al.
analyzed the complementarity of solar power and run-­ where PPV and P′PV are the actual and the rated power
of-­the-­river hydropower across different temporal scales output, respectively; RT is the irradiation on the device
using two indicators: the standard deviation of the energy surface; RSTC represents the solar radiation intensity under
balance and the theoretical storage required for balancing the standard test conditions, equivalent to 1000 W/m2;
generation and load, and demonstrated that at small αp is the temperature coefficients of power output of the
temporal scale (hourly), a high share of run-­of-­the-­river solar cell module, which is −0.35%/°C in this study; TC
power allows minimizing the energy balance variability is the actual temperature of the module, and TSTC is the

2 © 2018 The Authors. Energy Science & Engineering published by the Society of Chemical Industry and John Wiley & Sons Ltd.
F.-­F. Li et al. Multi-objective optimization of hydro-PV

temperature under the standard test conditions, equivalent the hydraulic and electric conditions and connections,
to 25°C. mainly includes:

Hydropower output model Water balance equations

Calculation of hydropower output Water balance equations shown in equation (4) imply

the hydraulic relations between neighboring time steps:
Hydropower station is a comprehensive engineering facility
that converts water power into electrical energy.
Hydropower station buildings concentrates water head of St+1 = St + Qtin × Δt − Qtout × Δt, (4)
natural flow, and transports the water flow to the turbine.
The water turbine and generator converts the concentrated where t is time index; S is the storage volume in reservoir;
hydro energy to electrical energy, which is then transmit- Qin is the inflow, and Δt is the length of the time period
ted into power grid through transformer, switch station, t.
and transmission lines. The basic calculation formula of
hydropower is shown in equation (2).
Water level-­storage curves of the reservoir
A nonlinear function f(x) is adopted to express the re-
PH = KQout ΔH (2)
lationship between storage S and water level L according
to the reservoir characteristic database:
where PH is the power output of the hydropower station;
K is the synthetic output coefficient of the power station,
whereas Qout is the discharge flow of the reservoir, and St = f (Lt ) ∪ Lt = f−1 (St ) (5)
ΔH is the net water head of the power station, which is
calculated by equation (3).

Tailwater elevation curves at reservoirs

ΔH = L − T (3)
Tailwater elevation T relates to the reservoir discharge,
which can be calculated based on discharge-­elevation curve
where L the reservoir water level, and T is the tailwater
g(x), as shown in equation (6):
elevation.
The calculation of hydropower output in equation (2)
involves multiple operational state equations representing Tt = g (Qtout ) (6)

Figure 1. Schematic diagram of a PV-­hydro hybrid power system.

© 2018 The Authors. Energy Science & Engineering published by the Society of Chemical Industry and John Wiley & Sons Ltd. 3
Multi-objective optimization of hydro-PV F.-­F. Li et al.

Optimization model for Hydro-­PV hybrid solutions are generated randomly, as shown in
power system equation (10):
Objective function [Qtactual × (1 − 𝛿)] ≤ Qtout ≤ [Qtactual × (1 + 𝛿)] (10)
Hydro-­PV projects often rely on existed hydropower sta-
tions, and the new photovoltaic power plants is added
into the system as the expansive capacity, as shown in
Figure 1. Energy production and energy consumption are Constraints
the key issues to be considered in power systems, which Most scheduling requirements of a reservoir such as flood
may be contradictory. In this study, maximizing the an- control, ice prevention, water supply, etc., can be realized
nual energy generation is set as the optimization objective by the control of water level and release, as shown in
Z1, as in equation (7). Another objective Z2 in equa- equations (11) and (12):
tion (8) is related to energy consumption, which aims at
minimizing the gap between the energy production and
energy consumption for the given system. The energy Ltmin ≤ Lt ≤ Ltmax (11)
generation is composed of hydropower and PV power, Qtmin ≤ Qtout ≤ Qtmax
resulting from equations (1) and (2). The average daily (12)
solar radiation intensity and temperature is used to cal-
culate the hydropower output scenarios in where Ltmin and Ltmax are the minimum and maximum
equation (1). water levels of the reservoir permitted at the t-­th time
(day or month) period; Qtmin and Qtmax are the minimum
� � and maximum water release of the reservoir permitted
Z1 =Max E Qtout �RtT , TtC =
∑N � � at the t-th time(day or month) period.
Max t=1 PtH (Qtout �Qtin ) + PPV (RtT , TtC ) × Δt, (7)

Optimization algorithm
∑N [ ]2 Pareto-­optimal solutions
Z2 = Min (PtH (Qtout |Qtin ) + PtPV (RtT , TtC ) − Dt , (8)
t=1
Without loss of generality, a multiobjective optimization prob-
lem can be defined as a vector minimization problem:
where E is the total generated energy, D is the energy
consumption.
MinimizeX∈Ω F (X) = (F1 (X), F2 (X), … , Fm (X))T , (13)
Decision variables
It can be seen from equations (7) and (8) that the only Subject to gi (X) ≥ 0 i = 1, 2, … , k, (14)
controllable variable in the hydro-­ PV power system is
the reservoir release, which determines the power genera-
tion of the hydropower station and thus affects the benefits hj (X) = 0 j = 1, 2, … , l, (15)
of the hybrid system. Hence, the reservoir release Qtout is
selected as the decision variable u in the optimization where X is a variable vector, Ω is a feasible solution
model, as shown in equation (9). space, and F1 (X), F2 (X), …, Fm (X) denote real valued
������
t ⃗
objective functions.
u = {Q out
} (9) Usually, the optimum for a multiobjective problem is
a solution set instead of a single solution, known as
Taking 1 month as the time step, there are 12 deci- Pareto-­optimal solutions. For any two points X1 and X2
sion variables for the annual optimization model. To in Ω, if the following conditions hold:
improve the calculation efficiency, the reservoir release
in the typical hydrological years deriving from frequency
curve Qtactual is used as the reference to construct the Fi (X1 ) ≤ Fi (X2 ) ∀i ∈ {1, … , m}, (16)
initial solution space. A certain percentage of the actual
discharge δ is added to the outflow trajectory to form
Fj (X1 ) ≤ Fj (X2 ) ∀j ∈ {1, … , m}, (17)
a band of discharge, in which the candidate initial

4 © 2018 The Authors. Energy Science & Engineering published by the Society of Chemical Industry and John Wiley & Sons Ltd.
F.-­F. Li et al. Multi-objective optimization of hydro-PV

then X1 is said to be superior to X2, vector X1 is at least as where X and Y are two individuals of population and
good as X2 with respect to all m objectives, and X1 is strictly xi and yi are objective functions that should be minimized
better than X2 with respect to at least one objective. If no other and M is number of objectives.
solution is superior to X1, then X1 is called a Pareto-­ optimal When the last front is under consideration, those con-
solution. The boundary consisting of the set of Pareto-­optimal figurations at a scarcely populated area which is far away
solutions is called a trade-­off surface, or a Pareto frontier. from the other solutions are selected to fill up the rest
of the positions. Such diversity is measured by crowding
distance as defined in equation (19).
NSGA-­II
Evolutionary algorithms have been proved to be effective M
∑ |fk+1 − fk−1 |
to deal with multiobjective problems (MOPs). The modi- D(k) =
j j

fmax − fmin (19)

fied version of Non-­dominated Sorting Genetic algorithm j=1 j j

(NSGA), NSGA-­ II, proposed by Deb et al. [17] is one where D(k) is crowding distance of individual k, f kj is
of the most efficient and commonly used evolutionary the j-­th objective function value of k-­th individual, fmax
j
algorithms with are profit from its simplicity, effectiveness, and fmin are the maximum and minimum values for the
j
and minimum user interaction [18]. j-­th objective function, respectively.
First, the initial population composed of individuals rep- Selection operation is used to make better-­performing
resented in equation (9) is generated randomly within the members have large survival probability. The individual
feasible domain shown in equation (10). The fitness of each with higher front number is selected, and in the same
individual, that is, the objective function value described in Pareto front number, the individual with bigger crowding
equations (7) and (8) is then worked out by various rela- distance is selected. After selection procedure, simulated
tionships shown from equations (1) to (6). Then the popula- binary crossover and polynomial mutation operation are
tion is layered and sorted to form multiple Pareto fronts employed to form children population. Elite strategy is
with different ranks based on the domination level of each adopted to let superior members in parent population
member determined by its own value of objective function. remain in children population. New parent population is
Such procedure is repeated until all solutions are set into formed by adding elite population members from the
fronts. The domination level is defined in equation (18): parents and children population based on the results of
nondomination sorting and crowding distance
X= {x1 , x2 , … , xM }
calculation.
Y= {y1 , y2 , … , yM } (18) A scheme of the proposed optimization model is given
X Dom Y⇔ ∀ i: Xi ≤ yi and ∃ j: xj ≤ yj , in Figure 2.

Figure 2. Proposed optimization model of hydro/PV hybrid system.

© 2018 The Authors. Energy Science & Engineering published by the Society of Chemical Industry and John Wiley & Sons Ltd. 5
Multi-objective optimization of hydro-PV F.-­F. Li et al.

Case Study joint operation, the full capacity of 850 MWp of the

PV power output can be compensated.
The Qinghai province is located in northwest of China In this study, the long-­time average annual value of the
with rich solar energy resources. The total annual radia- daily radiation and temperature at the Gonghe County is
tion of the Qinghai province is between 5800 and used to calculate the PV power output. Different hydrologic
7400 MJ/m2, more than 60% of which is the direct years from 2007 to 2014 are used to estimate the trade-­off
radiation [19]. The Longyangxia hydropower station on between the energy generation and the energy consump-
the Yellow River is located in the Gonghe County of tion, as well as the corresponding operational schemes for
the Qinghai province with the total installed capacity the dry years, normal years, and wet years, respectively.
of 4×320 MW. The storage capacity of the Longyangxia The NSGA-­II code in Matlab developed by Kanpur Genetic
reservoir is 14.700 billion m3, and it has good adjust- Algorithms Laboratory (KanGAL) [20] is used for the op-
ment ability with the regulating storage capacity of timization in this study.
19.35 billion m3. The PV power station is located on
the Talatan gobi desert in the Gonghe County with the
Results and Discussions
total installed capacity of 850 MWp. Figure 3 shows the
average daily solar radiation and temperature from the The mean annual discharge into the Longyangxia reservoir
year 1984 to 2004, which is taken to calculate the daily is 650 m3/s. Table 1 shows the modulus ratio coefficients
PV power output of the PV power station in this study. Kp defined in equation (20) of the inflow into the
The PV power is transmitted to the Longyangxia hy- Longyangxia reservoir from the year 2007 to 2014.
dropower station with a circuit of 330 kV, and the
compensatory power composed of hydropower and PV
power is transmitted to the power grid by five circuits. Kp = I∕Ī (20)
When two of the four hydropower units in the
Longyangxia hydropower station participate the comple- where I is the inflow into the reservoir, and Ī is the
mentary operation of hydro-­ PV hybrid power system, mean annual inflow into the reservoir.
75% of the 850 MWp of the PV power output can be According to the classification by [21], the year of 2012,
compensated within the scope of 0–640 MW; while three 2014, and 2008 are selected to represent the wet year,
or more sets of the hydropower units take part in the normal year, and dry year, respectively.

8 20

15
Solar radiation (kW/m2/day)

7
Temperature ( C)

10
6
5
5
0
4
Solar radiation -5

3 Temperature
-10

2 -15
1/1 2/1 3/1 4/1 5/1 6/1 7/1 8/1 9/1 10/1 11/1 12/1
Date

Figure 3. Average daily solar radiation and temperature from the year 1984 to 2004 at the Longyangxia PV power station.

Table 1. Analysis of the runoff into the Longyangxia reservoir.

Year 2007 2008 2009 2010 2011 2012 2013 2014

Annual runoff (m3/s) 587 521 754 586 622 858 590 636
Runoff modulus ratio 0.90 0.80 1.16 0.90 0.96 1.32 0.91 0.98
Classification Dry Dry Wet Dry Normal Wet Dry Normal

6 © 2018 The Authors. Energy Science & Engineering published by the Society of Chemical Industry and John Wiley & Sons Ltd.
F.-­F. Li et al. Multi-objective optimization of hydro-PV

In the NSGA-­II implementation, the population is set the energy generation, the variance between energy demand
to 200; the iteration stops after 500 times; the crossover and generation rises. Thus, maximizing energy generation
probability of Simulated Binary (SBX) Crossover [22] is is contradictory with minimizing the gap between energy
0.90; and the mutation probability of polynomial muta- demand and generation. Interestingly, the two objectives
tion [23] is 0.10. Premature convergence appears prominent almost keep a linear relationship with the same slope for
for genetic algorithms when decision variables have high different hydrologic years, although in the wet years, more
dimension and the searching space is large. It is quite energy could be generated. It implies that there exists a
common to get inferior solutions with small fitness values stable relationship between maximizing energy generation
or infeasible solutions during the early stage of the search- and minimizing the gap, which is determined by the
ing process [24]. In order to avoid the premature con- characteristics of the power system instead of the envi-
vergence, and take the power demand into account, the ronmental condition. In order to find out the optimal
historical reservoir discharge scenario is used as a bench- solution balancing the two objectives, the objectives are
mark to construct solution space to generate candidate normalized into [0, 1], as shown in Figure 4B–D. Since
individuals for GA. the Z1 needs to be as large as possible, whereas Z2 should
Figure 4 shows the Pareto Front of the selected years, be as close to 0 as possible, the idealist solution should
which represents the trade-­offs between energy generation be at the point (0, 1) in the coordinate system. The point
and the variance of energy demand and energy genera- on the Pareto Front nearest to (0, 1) is labeled in
tion, as well as the recommended optimal solutions. It Figure 4B–D, and taken as the recommend optimal solu-
can be shown from Figure 4A that with the increase in tion considering energy generation and the variance

Figure 4. Optimized results (A) Pareto Front, representing trade-­offs between energy generation and variance of energy demand and energy
generation of different hydrologic years; (B) recommended optimal solution of the year 2008, representing dry years; (C) recommended optimal
solution of the year 2014, representing normal years; (D) recommended optimal solution of the year 2012, representing wet years.

© 2018 The Authors. Energy Science & Engineering published by the Society of Chemical Industry and John Wiley & Sons Ltd. 7
Multi-objective optimization of hydro-PV F.-­F. Li et al.

Figure 5. Optimal operational scheme of (A) the year 2008, representing dry years; (B) the year 2014, representing normal years; (C) the year 2012,
representing wet years.

between energy generation and demand simultaneously. generation and consumption, the reservoir release process
The corresponding operation schemes, the reservoir release needs to be gentler, as the energy demand does not
scenarios, are presented in Figure 5, as well as those cor- fluctuate as violently as the reservoir inflow. The store
responding to the solutions with maximum energy gen- and delay effect of the reservoir on flow plays a more
eration and minimum variance, respectively. That is, important role when considering energy consumption.
Figure 5 shows the reservoir operation corresponding to As to the recommended operation considering both of
the denoted point and the two endpoints on the curve the two objectives, the release amount is basically in
in Figure 4. between. In nonflood season, the optimal operation is
The flood season lasts from June to September. more inclined to maximize energy generation, while it
Figure 5A shows that more water needs to be discharged concerns more about the energy consumption in flood
to maximize energy generation in dry year. But if the season. Another knowledge implied by the results is that
water is sufficient in normal or wet years as in Figure 5B the regulation function of the reservoir needs to be paid
and C, a better way to increase energy production is to more attention to in dry and normal years, when the
keep a high reservoir water level in nonflood season to operation schemes emphasizing different objectives dis-
improve power generation efficiency. Since the water level criminate greater.
has to be lowered to the flood control level, increasing The multiobjective model proposed in this study can be
reservoir discharge is always the primary choice in flood generalized to other hydro-­ PV hybrid power systems to
season to maximize energy generation for both dry and derive the instruction for the reservoir operation, and better
wet years. To minimize the variance between energy understand the characteristics of the power system.

8 © 2018 The Authors. Energy Science & Engineering published by the Society of Chemical Industry and John Wiley & Sons Ltd.
F.-­F. Li et al. Multi-objective optimization of hydro-PV

Acknowledgments remote villages in Cameroon. Renew. Energy

34:1445–1450.
This research was supported by National Key R&D Program 11. Beluco, A., P. K. de Souza, and A. Krenzinger. 2012. A
of China (2017YFC0403600, 2017YFC0403602), the Open method to evaluate the effect of complementarity in
Research Fund Program of State key Laboratory of time between hydro and solar energy on the
Hydroscience and Engineering, Tsinghua University (Grant performance of hybrid hydro PV generating plants.
No. sklhse-2016-B-03), and the Open Project of State Key Renew. Energy 45:24–30.
Laboratory of Plateau Ecology and Agriculture, Qinghai 12. Patsialis, T., I. Kougias, N. Kazakis, N. Theodossiou,
University.(Grant No. 2016-KF-03). and P. Droege. 2016. Supporting renewables’ penetration
in remote areas through the transformation of non-­
powered dams. Energies 9:1054.
Conflict of Interest
13. Francois, B., M. Borga, J. D. Creutin, B. Hingray, D.
None declared. Raynaud, and J. F. Sauterleute. 2016. Complementarity
between solar and hydro power: sensitivity study to
References climate characteristics in Northern-­Italy. Renew. Energy
86:543–553.
1. Ruhang, X. 2016. Characteristics and prospective of 14. Schmidt, J., R. Cancella, and A. O. Jr Pereira. 2016.
China’ s PV development route: based on data of world An optimal mix of solar PV, wind and hydro power
PV industry 2000–2010. Renew. Sustain. Energy Rev. for a low-­carbon electricity supply in Brazil. Renew.
56:1032–1043. Energy 85:137–147.
2. http://www.cgdc.com.cn/zhzx/1019892.jhtml. 15. Kougias, I., S. Szabó, F. Monforti-Ferrario, T. Huld,
3. Tang, Y. H., J. Yang. 2005. Status and expectation of and K. Bódis. 2016. A methodology for optimization of
photovoltaic technology. Renew. Energy 3:68–69. the complementarity between small-­hydropower plants
4. http://www.solarzoom.com/article-3847-1.html. and solar PV systems. Renew. Energy 87:1023–1030.
5. Zhang, P., T. Yang. 2015. Research on Longyangxia 16. http://www.nrel.gov/homer.
Hydro-­photovoltaic complementary operation mechanism. 17. Deb, K., A. Pratap, S. Agarwal, and T. Meyarivan. 2002.
J. N. China Univ. Water Resour. Electric Power A fast and elitist multiobjective genetic algorithm:
36:76–81. NSGA-­II. IEEE Trans. Evol. Comput. 6:182–197.
6. Wu, C. S., H. Liao, Z. L. Yang, Y. B. Wang, Y. C. Peng, 18. Niu, X., K. Liu, Y. Zhang, G. Xiao, and Y. Gong. 2017.
and H. H. Xu. 2009. Research on control strategies of Multiobjective optimization of multistage synchronous
small-hydro/PV hybrid power system. In Sustainable induction coilgun based on NSGA-­II. IEEE Trans.
Power Generation and Supply, 2009. SUPERGEN’09. Plasma Sci. 45:1622–1628.
International Conference on. IEEE. 19. http://www.gov.cn/guoqing/2013-03/26/content_5046173.
7. Meshram, S., G. Agnihotri, and S. Gupta. 2013. Power htm.
management strategy for active power sharing in hydro/ 20. http://www.iitk.ac.in/kangal/codes.shtml.
PV/battery hybrid energy system. Chin. J. Eng. 2013: 21. Ba Cunsheng, Z. J. 2003. Analysis of Longyangxia
1–7. reservoir’s entering radial flow character and period.
8. Kenfack, J., F. P. Neirac, T. T. Tatietse, D. Mayer, M. Qinghai Electric Power 4:4–6.
Fogue, and A. Lejeune. 2009. Microhydro-­PV-­hybrid 22. Deb, K., and R. B. Agrawal. 1995. Simulated binary
system: sizing a small hydro-­PV-­hybrid system for rural crossover for continuous search space. Complex Syst.
electrification in developing countries. Renew. Energy 9:115–148.
34:2259–2263. 23. Deb, K., and M. Goyal. 1996. A combined genetic
9. Bekele, G., and G. Tadesse. 2012. Feasibility study of adaptive search (GeneAS) for engineering design.
small Hydro/PV/Wind hybrid system for off-­grid rural Comput. Sci. Inform. 26:30–45.
electrification in Ethiopia. Appl. Energy 97:5–15. 24. Chang, L.-C., F.-J. Chang, K.-W. Wang, and S.-Y. Dai.
10. Nfah, E., and J. Ngundam. 2009. Feasibility of pico-­ 2010. Constrained genetic algorithms for optimizing
hydro and photovoltaic hybrid power systems for multi-­use reservoir operation. J. Hydrol. 390:66–74.

© 2018 The Authors. Energy Science & Engineering published by the Society of Chemical Industry and John Wiley & Sons Ltd. 9