CHAPTER 1

SCALAR CONTROL OF INDUCTION MOTORS

AIM OF THE LECTURE: The main objectives of this first lesson is to developp the principle equations and
schemes, which allow the closed loop speed tracking of an induction motor, while keeping the motor to
operate under constant air-gap flux. By the way, this type of control is usually known as:
1- Scalar Control
2- Constant air-gap flux control
3- Constant Torque Control

One essential Question needs to be tackled: On which parameter we have to set in this kind of control?

As you know guys, the rotational speed of the IM can be put as:
1 − 𝑔 𝑓ൗ
𝑛= 𝑝
1- Speed control via pair-poles number Discrete speed control
2- Slip control Continuous speed variation but limited to specified applications
(pumping,…….., where the load torque is generally quadratic).
3- Frequency control Continuous, with wide intervalle of speed control (That’s it)
• In steady state, the equivalent scheme of an induction motor can be made as depicted in the following
figure.

The electromotive force E1 of such machine can be

modeled via the following equation:

𝐸1 = 4.44. 𝐾. 𝑛. 𝑓. 𝜑
Where:
K denotes the winding coefficient.
n is the number of spires.
f denotes the stator frequency, and 𝜑 is the airgap flux

The functioning under a constant airgap flux 𝜑 = 𝜑𝑟𝑎𝑡𝑒𝑑 , needs only to keep the caution:

𝐸1
ൗ𝑓 = 𝐶𝑜𝑛𝑠𝑡𝑎𝑛𝑡
While operating in considerable speed range, the statoric voltage loss 𝑅𝑠 + 𝑗𝑋1 . 𝐼𝑠 can be omitted,
and the following equation holds:
𝐸1 ≅ 𝑉𝑠
Which implies as a consequence the following:
𝑉 𝑉 𝑉 𝑉 𝑉
Functioning under 𝜑=Const, → 𝑠ൗ𝑓 = 𝐶𝑜𝑛𝑠𝑡, 1ൗ𝑓1 = 2ൗ𝑓2= 3ൗ𝑓3= 4ൗ𝑓4 = 220Τ50 = 4.44
 As a consequence, the mechanical characteristics of such control are shown in the following
figure:

As can be seen, the speed is widely changed from standstill to rated value. Despite of these good
performances, one problem occurs: the maximum Torque (breackdown torque) can’t be kept in small
values of frequency: What’s the problem guys ?
 This is mainly caused by the stator voltage loss 𝑅𝑠 + 𝑗𝑋1 . 𝐼𝑠 which become considerable:
At small frequency values, 𝑉𝑠 = 4.44. 𝑓 = 𝑅𝑠 + 𝑗𝑋1 . 𝐼𝑠 + 𝐸1
Since 𝑉𝑠 is somewat small, and the voltage loss increases, the back-Emf decreases, which leads to
a dramastic decrease of the airgap-flux 𝜑. As a result the breackdown torque 𝑇𝑚𝑎𝑥 is no longer
kept constant.
Hey, Audiance, is there any problem with that ?
Unfortunately yes,

Since 𝑇𝑚𝑎𝑥 decreases, the overload margin, which equals: 𝑇𝑚𝑎𝑥ൗ𝑇𝑟𝑎𝑡𝑒𝑑 decreases significantly,
and the motor can face difficulties to handle considerable load torque values.
 What’s the solution so ?

The solution is to add a voltage quantity 𝑉0 to the stator voltage

Consequently, the voltage law becomes:

𝑽𝑺 = 𝑲. 𝒇 + 𝑽𝟎
The associated V-f curve which summeries the operation zones of such control is ploted in
the following curve:
 In this figure, two zones are depicted:
1- With frequencies below the rated values (𝑓 ≤ 𝑓𝑟𝑎𝑡𝑒𝑑 , 𝜑 = 𝜑𝑟𝑎𝑡𝑒𝑑 and 𝑽𝑺 = 𝑲. 𝒇 + 𝑽𝟎 ): The
motor operates under constant flux and torque.

2- Beyond 𝑓𝑟𝑎𝑡𝑒𝑑 , the voltage is kept constant (𝑉𝑠 = 𝑉𝑟𝑎𝑡𝑒𝑑 ), and the caution: 𝜑 = K. 𝑉𝑠ൗ𝑓
decreases. We call that: Flux weakening. The motor operates in constant power mode.

An additional Remark can be provided:

With constant load torque operation, the slip frequency (𝑤𝑠𝑙 = 𝑤𝑠 − 𝑤𝑟 ) is kept constant
This is in fact the key to perform a closed loop scalar control.

An essential question must be answered:

on which parameter we have to adjust in speed closed loop?

The mechanical equation of any elctrical motor is given by:

𝑑𝜔
𝐽 + 𝑓𝜔 = 𝐶𝑒 − 𝐶𝑟 , where 𝐶𝑒 and 𝐶𝑟 denote the electromagnetic and load torque respectively
𝑑𝑡

The answer of the above question is: One has to adjust the electromagnetic torque 𝐶𝑒 or its image
 The elctromagnetic torque of an IM can be expressed by the following equation: 𝐶𝑒 = 𝐾1 . 𝜑 2 . 𝑤𝑠𝑙
Under constant flux operation, 𝐶𝑒 is proportionnal to 𝑤𝑠𝑙 , which is its image.

• Closed loop speed control: The voltage approach:

In this type of control, the approaching control of the Voltage source inverter (VSI) is performed in voltage
approach, through the use of a Sine-triangular PWM concept.
The purpose is to ensure a reference speed 𝝎𝒓𝒆𝒇 tracking, while maintaining a constant air gap flux operation

Integrate a Proportional-Integral controller to keep a speed tracking.

According to the controller output 𝑤𝑠𝑙 construct the voltage equation 𝑽𝑺 = 𝑲. 𝒇 + 𝑽𝟎 , and the
autopilot equation: 𝑤𝑠 = 𝑝. 𝑤𝑟 + 𝑤𝑠𝑙

Construct the three reference voltage equations: 𝑽𝑺1 = 2. 𝐾. 𝑓 + 𝑉0 . sin 𝑤𝑠 . 𝑡 ,

……..
Use the sinus-triangular PWM strategy to control the six switches of the VS Inverter.
Speed Controller: It is in fact a PI controller, with a Proportional action ‘P’, and an Integral action ‘I’
The input of this bloc is the tracking error: 𝑒 = 𝑤𝑟𝑒𝑓 − 𝑤𝑟 , and the output is the slip frequency 𝑤𝑠𝑙
𝑑𝑒(𝑡)ൗ
𝑤𝑠𝑙 = 𝐾𝑝 . 𝑒 𝑡 + 𝐾𝑖 𝑑𝑡
Where 𝐾𝑝 and 𝐾𝑖 are the controller’s gains to be identified.
𝐾𝑝 𝑎𝑛𝑑 𝐾𝑖 selection: Various temporal and frequencial approaches exist in literature to select
the controller’s gains.

The Ziegler-Nichols PID Tuning Method: The above equation of the controller can be re-
1
written as: 𝐶(𝑠) = 𝐾𝑝 (1 + + 𝑇𝑑 𝑠)
𝑇𝑖 𝑠
- The first step is to set 𝑇𝑑 = 0 and to set 𝑇𝑖 = ∞. That is, we only leave the proportional
action. In that case, the controller is: 𝐶(𝑠) = 𝐾𝑝 . The closed loop becomes:

- The second step: Starting from zero, we keep on increasing 𝐾𝑝 until stable oscillation is
observed. We read the period 𝑇𝑜𝑠𝑐𝑖 of such a stable oscillation. The period 𝑇𝑜𝑠𝑐𝑖 of stable
oscillations is called the critical period or the ultimate period. The period 𝑇𝑜𝑠𝑐𝑖 is the
input parameter for designing the PID controller.
- The third step: The proportional gain 𝐾𝑝 ∗ that produces the stable oscillation is called the
critical gain or the ultimate gain.
𝐾𝑝 = 0.6 ∗ 𝐾𝑝 ∗ , 𝑇𝑖 = 0.5 ∗ 𝑇𝑜𝑠𝑐𝑖 , 𝑇𝑑 = 0.125 ∗ 𝑇𝑜𝑠𝑐𝑖
Simulation Results:
• Current Control Strategy:
• Another Point of view exists to ensure the same performances, where the IM is controlled in current, and
not in voltage.
• Let’s recall the equivalent scheme of the IM, presented in the second slide.

By neglecting the core resistance 𝑅𝑐 , the air-gap flux can be

expressed by the following:
𝜑 = 𝐿𝑚 . 𝐼0
Where 𝐼0 is the magnetizing current. Obviously, to keep 𝜑 =
𝐶𝑜𝑛𝑠 implies that 𝐼0 is also constant
Accordingly, a close relationship exists between the
effective value of stator current 𝐼𝑠 and the slip frequency
𝑤𝑠𝑙 .

𝜔𝑠𝑙. 𝐿𝑟 2
∅𝑠_𝑟𝑎𝑡𝑒𝑑 1+
𝑅𝑟
𝐼𝑠 = ,
𝐿𝑠 𝜎. 𝜔𝑠𝑙 . 𝐿𝑟 2
1+
𝐿𝑟

𝐿2𝑚 Is the leackage coefficient

And 𝜎 =1− ൗ𝐿𝑠 .𝐿𝑟
Control Steps:

The main differences between this type of control and the previous voltage control technique are:
- Instead of writing the famous voltage equation: 𝑉 = 𝑓 𝑤𝑠𝑙 , write the current equation 𝐼𝑠 = 𝑔 𝑤𝑠𝑙 , or
the analogue lookup table.
- Construct three reference currents instead of three reference voltage: 𝐼𝑺1 = 2. 𝐼𝑠1 . sin 𝑤𝑠 . 𝑡 ,
……..
- Perform a current control of the three phase VS inverter, through the Hysterisis controller. This latter
allows the stator currents to track their set points according a predifined hysterisis band ∆.
 By choosing a large value of the deviation current ∆= 𝐼𝑟𝑒𝑓 − 𝐼𝑠 , the performances are poor, but the
switches heating is moderate.
 A small value, increases considerably the performances, but the constraints on the switches safety
increase as a consequance. As a result a trade off must be done between performances and switches
safety.
Current Based Scalar Control of IM
 Photovoltaic Pumping System
• In this section, we exhibit the association of a PV array with the above electric drive, where the application is
widely realized worldwide: The PV pumping system. As described in the following synoptic scheme: A PV
array, considered as the sole source, supplies the motor group via a boost DC-DC and a VS inverter. A
centrifugal pump sucks up the water from a first tank, and pushes it back, with high pressure to a second
tank.
• The DC-DC converter, acts as an impedance adapts, and extracts the maximum PV power permanently, via
an MPPT algorithm.
• A voltage based scalar control of the IM is adopted.

MPPT Control Scalar Control

 System Modeling

1- PV array modeling:
The PV generator is represented by an implicit I-V characteristic, where the maximum power is located at the
Knee of this curve.
𝐼𝑝𝑣 = 𝐼𝑠𝑐 − 𝐼0 . (𝑒𝑥𝑝 𝑉𝑝𝑣 + 𝑅𝑠 . 𝐼𝑝𝑣 − 1)
This equation is mainly dependant on the variation of the solar irradiation E and the cell temperature.
 2- Induction motor medeling: The IM is represented by the following state space model :

• The load torque of the centrifugal pump is represented by a quadratic equation: 𝐶𝑟 = 𝐾2 . 𝜔2

• So, if the sytem is lossless, the generated PV power is completely converted in mechanical power; which
implies that: 𝑃𝑝𝑣 = 𝐾2 . 𝜔3
• 3- Hydraulic model: The centrifugal manometric height 𝐻𝑡 is represented by a quadratic equation as
function of the water flowrate 𝑄, and the motor speed 𝜔:
𝐻𝑡 = 𝑎0 . 𝜔2 −𝑎1 . 𝜔. 𝑄 − 𝑎2 . 𝑄2
• The functioning point is located at the intersection
of the above equation and the load (discharge pipe):
𝐻𝑡 = 𝑏0 + 𝑏1 . 𝑄 2
 Control of the PV system:

1- PV Side Control: On this side, a Perturb & Observe (P&O) algorithm is used to permanently extract the
maximum PV power, despite the solar irradiance changing. The idea is to perturb the system with an incremetal
duty cycle ∆𝑫 , and observe the power positive or negative ∆𝑷𝑷𝑽 .
2-Motor Side Control:

The Induction motor is controlled by the voltage scalar control, and where the reference speed is obtained
by the following formulation:
𝝎𝒓𝒆𝒇 = 𝝎𝒓𝒆𝒇_𝟏 + 𝝎𝒓𝒆𝒇_𝟐
where :
𝝎𝒓𝒆𝒇_𝟏 is obtained from the DC link voltage 𝑽𝒅𝒄 regulation:
𝝎𝒓𝒆𝒇_𝟏 = 𝑷𝑰 (𝑽𝒅𝒄_𝒓𝒆𝒇 − 𝑽𝒅𝒄 )

𝟑 𝑷𝒑𝒗
and 𝝎𝒓𝒆𝒇_𝟐 = ൗ𝑲𝟐
 (P&O)
MPPT Control

Scalar Control

Control of the PV pumping System