ADVANCE ELECTRIC MACHINE

Equivalent Circuit and

Methods to Determine the
System Parameter of IM
PRESENTED BY :

M UH A MMA D R I ZA N I R U SL I ( 1 0 2 0 17 1 01 2 )
LECTURER :
Dr. Ir. Era Purwanto, M. Eng.
NIP. 19610601.198701.1.001

Program Studi
Magister Terapan Teknik Elektro
Program Pascasarjana Teknologi Rekayasa
Departemen Teknik Elektro
Politeknik Elektronika Negeri Surabaya
2018
Equivalent Circuits and Methods to Determine the
System Parameter

- Equivalent circuits
o Starting point for control design based on the representation of the
physical quantities will be a very useful tool
o Equivalent circuit using symmetrical conditions assumption and need
accuracy of equivalent circuit parameter
o Classical method to obtain equivalent circuit parameter are not
practicable. Possible to automated computation
 Equivalent Circuits of IM
- T equivalent circuits
o Voltage and flux linkage equations
𝑢𝑠 = 𝑅𝑠 𝑖𝑠 + 𝜑𝑠
𝑢𝑟 = 𝑅𝑟𝑟 + 𝑖𝑟𝑟 + 𝜑𝑟𝑟 − 𝑗𝜔𝜑𝑟𝑟
𝑟

𝜑𝑠 = 𝐿𝑠 𝑖𝑠 + 𝐿𝑚 𝑖𝑟
𝜑𝑟𝑟 = 𝐿𝑟𝑟 𝑖𝑟𝑟 + 𝐿𝑟𝑚 𝑖𝑠𝑟
o T model : ideal transformer model with an additional secondary
𝑢
(rotor-sided). Transfer ratio 𝑡𝑟 = 𝑢𝑟𝑁 Figure 1. Transformer equivalent circuit of IM
𝑠𝑁
o For the transformation of the rotor quantities to the stator side
by using 𝑡𝑟 .
𝑢𝑟𝑟 𝜑 𝑟
𝑟
𝑢𝑟 = ; 𝑖𝑟 = 𝑡𝑟 𝑖𝑟𝑟 ; 𝜑𝑟 =
𝑡𝑟 𝑡𝑟
𝑅′𝑟 2
𝐿𝑟
𝑅𝑟 = 2 ; 𝐿𝑟 = 2 ; 𝑖𝜇 = 𝑖𝑠 + 𝑖𝑟
𝑡𝑟 𝑡𝑟 𝑟
 Equivalent Circuits of IM (1)
- T equivalent circuits
o From voltage and flux linkage equations
𝑢𝑠 = 𝑅𝑠 𝑖𝑠 + 𝜑𝑠
𝑢𝑟𝑟 = 𝑅𝑟𝑟 + 𝑖𝑟𝑟 + 𝜑𝑟𝑟 − 𝑗𝜔𝜑𝑟𝑟
𝜑𝑠 = 𝐿𝑠 𝑖𝑠 + 𝐿𝑚 𝑖𝑟
𝜑𝑟𝑟 = 𝐿𝑟𝑟 𝑖𝑟𝑟 + 𝐿𝑟𝑚 𝑖𝑠𝑟
o Equations of stator and rotor voltage
𝑑𝑖𝑠 𝑑𝑖𝜇
𝑢𝑠 = 𝑅𝑠 𝑖𝑠 + 𝐿𝑠𝜎 + 𝐿𝑚
𝑑𝑡 𝑑𝑡
𝑑𝑖𝑟 𝑑𝑖 𝜇
0 = 𝑅𝑟 𝑖𝑟 + 𝐿𝑠𝜎 + 𝐿𝑚 − 𝑗𝜔𝜑𝑠
𝑑𝑡 𝑑𝑡 Figure 2. T equivalent circuit of IM (a) non-stationary, (b)
o Transform into Laplace domain with operation (𝑠 → 𝑗𝜔𝑠 ) with stationary
𝜔𝑠 −𝜔
𝑠= 𝜔
𝑠
𝑢𝑠 = 𝑅𝑠 𝑖𝑠 + 𝑗𝜔𝑠 (𝐿𝑠𝜎 𝑖𝑠 + 𝐿𝑚 𝑖𝜇 )
𝑅𝑟
0= 𝑖 + 𝑗𝜔𝑠 (𝐿𝑠𝜎 𝑖𝑟 + 𝐿𝑚 𝑖𝜇 )
𝑠 𝑟
 Equivalent Circuits of IM (2)
- Inverse T equivalent circuits
𝜑𝑟 𝐿𝑚 𝐿2𝑚
o With 𝑖𝑚 = = 𝑖𝑠 + 𝑖 and leakage factor 𝜎 = 1 −𝐿 𝐿 new
𝐿𝑚 𝐿𝑟 𝑟 𝑠 𝑟
voltage equations become
𝑑𝑖𝑠 𝑑𝑖𝑚
𝑢𝑠 = 𝑅𝑠 𝑖𝑠 + 𝜎𝐿𝑠 + (1 − 𝜎)𝐿𝑠
𝑑𝑡 𝑑𝑡
2
𝐿𝑚 𝐿𝑚 𝑑𝑖𝑚
0 = 2 𝑅𝑟 ( 𝑖𝑟 ) + (1 − 𝜎)𝐿𝑠 − 𝑗𝜔𝜑𝑠 (1 − 𝜎)𝐿𝑠 𝑖𝑚
𝐿𝑟 𝐿 𝑟 𝑑𝑡

Figure 3. T equivalent circuit of IM (a) non-stationary, (b)

stationary
 Equivalent Circuits of IM (3)
- T equivalent circuits
o Transform leakage inductance into rotor side
𝜑 𝐿
𝑖𝑚𝑠 = 𝐿 𝑠 = 𝑖𝑠 + 𝐿𝑚 𝑖𝑟
𝑠 𝑠
is introduced analogously to inverse T equivalent circuit.
Substitution of 𝑖𝜇 to T equivalent circuit represented in the non-
stationary and stationary form will be obtained
𝑑𝑖𝑚𝑠
𝑢𝑠 = 𝑅𝑠 𝑖𝑠 + 𝐿𝑠
𝑑𝑡
2
𝐿𝑠 𝐿𝑚 𝜎𝐿𝑠 𝐿𝑚 𝑑𝑖𝑟 𝑑𝑖𝑚𝑠 𝐿𝑚
0 = 2 𝑅𝑟 ( 𝑖𝑟 ) + + 𝐿𝑠 − 𝑗𝜔 𝜑
𝐿𝑟 𝐿𝑟 1 − 𝜎 𝐿𝑠 𝑑𝑡 𝑑𝑡 𝐿𝑠 𝑟
The stator inductance now becomes the cross or magnetization
inductance, and the stator flux linkage assumes the role of the
main flux linkage. Therefore, T equivalent circuit is particularly Figure 4. T equivalent circuit of IM (a) non-stationary, (b)
stationary
suitable for treatment of stator flux oriented control methods.
Modelling of the Nonlinearities of the IM
o Many controls task the assumption of constant and state
independent machine parameters (symmetrical conditions and
sinusoidal winding). Therefore, nonlinearities which are
significant for different operating states into machine and
equivalent circuits.
o Nonlinearities following the physical conditions : magnetic
saturation, iron losses, and current displacement. Transform
leakage inductance into rotor side.
o In mathematical sense, nonlinear relations are indicated by the
fact that the superposition principle is not valid. Therefore,
nonlinearities is not permitted to be spoken strictly.
o Advantages of modeling nonlinearities : nonlinearities are
describable as state dependent parameters.
 Modelling of the Nonlinearities of the IM (1)
Iron Losses
o Iron losses consist of eddy-current losses and hysteresis losses.
o Iron losses : rotor and stator. Iron losses in rotor can be neglected because the
rotor frequency smaller than stator frequency
o Hysteresis loss : produced by the flux reversal energy consumed due to
sinusoidal with time varying iron magnetization. Depend on the effective
permeability.
o Eddy current losses proportional to the square of the voltage induced in the
iron
o Modeling is made difficult because of the effect of eddy-currents and
hysteresis from sinusoidal magnetization are overlapping in a not exactly
determinable way
o The hysteresis losses depend on the effective permeability and therefore on
the magnetic saturation on the instantaneous flux amplitude. They disappear
as soon as the area of the magnetic saturation is left (the upper field Figure 5. Eddy current and hysteresis loss
weakening area)
 Modelling of the Nonlinearities of the IM (2)
o Ideal model according to Murata 1990 takes into account hysteresis and eddy-current
parts by respectively constant factors 𝑘ℎ𝑦 and 𝑘𝑤
3
𝑃𝑣,𝑓𝑒 = 2 [𝑘ℎ𝑦 (𝜔𝑠 + 𝜔𝑟 ) + 𝑘𝑤 𝜔𝑠 2 + 𝜔𝑟 2 ]|𝜑𝜇 |2 𝑤𝑖𝑡ℎ 𝜑𝜇 = 𝐿𝑚 𝑖𝜇
o With 𝑠 = (𝜔𝑠 − 𝜔)/𝜔𝑠 general equation for the iron losses
3 𝜔𝑠 |𝜑𝜇 |2
𝑃𝑣,𝑓𝑒 =
2 𝑅𝑓𝑒
o Iron loss resistance 𝑅𝑓𝑒 can be introduced
1
𝑃𝑣,𝑓𝑒 =
2
𝑘ℎ𝑦
𝑘𝑤 1 + 𝑠 + 𝜔 (1 + 𝑠)
𝑠
Figure 6. Iron losses and iron loss resistance
o Fig 6 shows no-load measurement at an inverter-fed and external driven motor so that
influence of the friction losses is eliminated
o Iron loss power is dominated by the hysteresis losses rising nearly linearly in the basic
speed range
o Solution : with field-weakening setting in, at first a strong drop can be observed because
of the flux reduction.
 Modelling of the Nonlinearities of the IM (3)
o The eddy current losses dominate in the upper field-weakening area. In
addition, the inverter dependent eddy-current losses decrease strongly at the
maximum voltage
o So that 𝑘𝑤 used in constant flux area and constant voltage range. (Figure 6
dotted lines)
o The possibility simplified model which only contains a linear relation between
loss resistance and stator frequency . For this model, the loss resistance at
nominal frequency must be determined by measurement.
𝜔𝑠
𝑅𝑓𝑒 = 𝑅𝑓𝑒𝑁
𝜔𝑠𝑁
Figure 6. Iron losses and iron loss resistance
 Modelling of the Nonlinearities of the IM (4)
o A comfortable inclusion in the equation system of the IM will be obtained if
the iron loss resistance in the inverse T equivalent circuit can be represented
by a parallel resistance in the stator circuit. It can be shown in Figure 7.
o The modified system is obtained to

Figure 7. Extended inverse T equivalent circuit with iron losses

 Modelling of the Nonlinearities of the IM (5)
Current and Field Displacement
o The current displacement leads to a frequency dependent increase of the
resistance values, and the field displacement to a reduction of the leakage
inductances.
o As a consequence the current harmonics produce higher losses. Because the
harmonic spectrum of fast switching inverters with sine modulation is orders of
magnitude above the fundamental wave, its significance for control related
parameters remains small.
o The consequences of the fundamental dependent current displacement,
however, must be investigated for the modeling of the machine.
 Modelling of the Nonlinearities of the IM (6)
Current and Field Displacement
o In stator windings of induction machines, fundamental wave dependent current
displacement effects can usually be neglected since they are intentionally suppressed
by a number of constructive measures.
o An exception would merely be the big machines with accordingly large winding
diameters at high frequencies. For the bars of the rotor squirrel-cage such a
neglection is not possible from the outset because of the large bar heights and
diameters, except for rotors with intentionally current displacement free construction.
o In the normal (stationary) operation current displacement effects do not play a Figure 8. Extended equivalent circuit with current displacement
in rotor
considerable role, however, because of the low rotor frequency (slip).
o This turns different in special non-stationary operation modes with high slip
frequencies, where the current displacement is used with purpose to increase the
resistance, or if the input quantities are controlled differently to the normal
operation.
o In a field-orientated control system however, also at start-up extreme slip values will
not appear due to the current being controlled with defined amplitude and slip.
 Modelling of the Nonlinearities of the IM (7)
Magnetic Saturation
o At first the magnetic saturation has the consequence that the value of the
inductances is a nonlinear function of the amplitude of the actual flux linkage.
o In addition, a general analysis of the saturated induction machine must take
into account that the spatial distribution of the saturation depends on the
current direction of the accompanying flux vector.
o This has the consequence that in the right-angled coordinate system the
inductances assigned to the coordinate axes assume different values in the
dynamic case, and mutual couplings appear (Vas 1990)
o These depend on the sine of the angle between the main flux vector and the
reference axis (real axis) of the used coordinate system.
o The main field saturation has essential significance for the dynamic behavior of
the machine, primarily in the field weakening and at great torques.
 Parameter Estimation from Name Plate Data
o Deviations from 50 to 100 % depending on the parameter in question
have to be taken into account, because:
• the manufacturer’s information may be partly unreliable, and the actual
motor parameters are subject to spreads,
• the name plate data refer to a certain working point (the nominal working
point),
• not all parameters of the equivalent circuit can be directly set into a
physical relation to the usual name plate data
o The usual name plate data are:
• Nominal power PN [kW]
• (Line-to-line) nominal voltage UN [V]
• Nominal current IN [A]
• Nominal frequency fN [Hz]
• Nominal speed nN [rpm]
• (Nominal) power factor cos ϕ
 Parameter Estimation from Name Plate Data (1)
Calculation for IM with Power Factor cos ϕ
o Deviations from 50 to 100 % depending on the parameter in question
have to be taken into account, because:
• the manufacturer’s information may be partly unreliable, and the actual
motor parameters are subject to spreads,
• the name plate data refer to a certain working point (the nominal working
point),
• not all parameters of the equivalent circuit can be directly set into a
physical relation to the usual name plate data
o The usual name plate data are:
• Nominal power PN [kW]
• (Line-to-line) nominal voltage UN [V]
• Nominal current IN [A]
• Nominal frequency fN [Hz]
• Nominal speed nN [rpm]
• (Nominal) power factor cos u
 Parameter Estimation from Name Plate Data (2)
Calculation for IM without Power Factor cos ϕ
o Deviations from 50 to 100 % depending on the parameter in question
have to be taken into account, because:
• the manufacturer’s information may be partly unreliable, and the actual
motor parameters are subject to spreads,
• the name plate data refer to a certain working point (the nominal working
point),
• not all parameters of the equivalent circuit can be directly set into a
physical relation to the usual name plate data
o The usual name plate data are:
• Nominal power PN [kW]
• (Line-to-line) nominal voltage UN [V]
• Nominal current IN [A]
• Nominal frequency fN [Hz]
• Nominal speed nN [rpm]
• (Nominal) power factor cos u
Thank You