Next Article in Journal
Circular Economy in Practice: Building a Simple Greenhouse from Recycled Plastic
Previous Article in Journal
Recent Advancements in the Tribological Modelling of Rough Interfaces
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Novel Suppression Method for Low-Order Harmonics Causing Resonance of Induction Motor

1
College of Electrical Engineering, Southwest Minzu University, Chengdu 610225, China
2
Institute of National Productivity Development, Southwest Minzu University, Chengdu 610225, China
*
Author to whom correspondence should be addressed.
Machines 2022, 10(12), 1206; https://doi.org/10.3390/machines10121206
Submission received: 3 November 2022 / Revised: 9 December 2022 / Accepted: 12 December 2022 / Published: 13 December 2022
(This article belongs to the Topic Designs and Drive Control of Electromechanical Machines)

Abstract

:
In the motor drive system of electric vehicles, there are some nonlinear factors, such as the deadtime and the conduction voltage drop of switching devices, which will generate low-order harmonics of the stator current and the torque ripple. The frequency of the harmonic may coincide with the natural frequency of the motor, so resonance may occur on the motor drive system. To reduce the noise caused by motor resonance, the characteristics of harmonic distortion caused by the deadtime, and the conduction voltage drop of the switching device, are analyzed firstly. Then, a motor vector control strategy with specific low order is proposed. The sixth-order harmonic resonance controller in d-q axis is introduced into the control loop, and the parameter designing principle of the controller is also presented. Without affecting the control performance of the current loop, the sixth-order harmonic of the stator current near the natural frequency can be significantly suppressed. Finally, the simulation and the experiment are carried out to certify the correctness and effectiveness of the proposed harmonic suppression method.

1. Introduction

The development of electric vehicles is an effective means to solve the problems of energy and environment [1], but serious vibration and noise exist in the motor drive system of the electric vehicle [2]. The output voltage of the inverter is a square wave instead of a sine wave [3], and contains abundant high-order harmonics around the switching frequency and its integer multiples [4]. These higher-order harmonics will lead to serious electromagnetic vibration and noise, and the life and comfort of the electric vehicle are subsequently reduced [5,6]. To reduce the power of the harmonics near the switching frequency, some modulation strategies, such as the random carrier frequency [7] and periodic carrier frequency [8], are proposed. The above methods can disperse the main harmonics spectrum into a wider frequency band, a good suppression effect on the high-frequency noise of the motor caused by the harmonics of stator currents.
Ideally, there is no low-order harmonic in the inverter output voltage [9], but due to nonlinear factors such as the deadtime [10], the conduction voltage drop of switching devices and body diode [11], and the snubber capacitor [12], the lower order harmonics of the stator current will be generated. The natural frequency of the induction motor (IM) for electric vehicles is low, and the operating speed range of IM is wide, so the frequencies of these low-order harmonics may coincide with the natural frequencies of the motor, and the resonance may occur on the motor drive system of the electric vehicle [13]. In reference [14], a reasonable slot fit is selected to weaken the electromagnetic force wave related to electromagnetic vibration, and the vibration peak value of the IM at the natural frequency can be reduced. In reference [15], a Y-type capacitor is introduced at the DC terminal of the inverter; it is particularly effective to reduce the radiation emission of the vehicle in the above frequency range, and a simplified vehicle common-mode interference model is established. Compared to the above two methods for the structural improvements of the inverter or motor, it is more economical and practical to improve the modulation and control algorithm. Compensation based on pulse time is the most conventional and commonly used low-order harmonic suppression method. According to the polarity of the stator current, the voltage error caused by the deadtime and the conduction voltage drop of switching devices is converted into the duration of PWM pulse, and is subtracted in a switching period [16]. However, the compensation effect of this method is not very obvious. A modulation method without deadtime is proposed in [17]; according to the current polarity, only one switching tube of a phase bridge arm remains in the active state, and the other tube remains OFF state. The freewheeling is completed by the body diode, but the current harmonics caused by the conduction voltage drop of switching devices cannot be suppressed. In reference [18], a feedforward compensation of the average error on the command value is proposed, which can reduce low-order harmonic distortion, but the suppression effect is limited and depends on the sampling accuracy of the current polarity. Reference [19] proposes a scheme for suppressing the 5th and 7th harmonics based on a PI controller, this scheme needs to convert the 5th and 7th harmonic components of the stator current to d-q axis, and then four PI controllers are adopted to obtain the compensation amount of the 5th and 7th harmonics. Although the suppression effect is superior to other methods, the implementation process of this method is a little complicated. In order to suppress the low-order harmonics of the stator current, reference [16] optimizes the current PI controller parameters, but this scheme will reduce the dynamic response speed of the control system. Reference [13] proposes a current spectrum shaping control scheme in the natural frequency range based on a band-pass filter, but the optimal parameters design principle of the band-pass filter is not considered. In reference [20], a frequency adaptive selective harmonic control implemented in Z-plane is proposed for grid-connected inverters, and the fundamental frequency is fixed at 50 Hz. However, the speed range of the motor drive system is wide, the applicability of this method needs further verification. In reference [21], the resonance frequency is adjusted adaptively, however, whether the resonant controller with the same gain and different frequencies will affect the control performance of the current loop needs to be verified. Meanwhile, the calculation expression of the fundamental angle frequency contains IM parameters, rotor flux, stator current in q-axis and speed, so the motor parameters variation and the sampling accuracy will easily lead to the calculation error of the fundamental frequency.
In this paper, the characteristics of harmonic distortion caused by deadtime, and the conduction voltage drop of switching devices, are analyzed first. Then, the vector control scheme is improved, a sixth-order harmonic proportional resonance (PR) controller in d-q axis is introduced in the control loop, and the parameters of the PR controller are optimally tuned in theory. Finally, the simulation and experimental validations of the proposed method are carried out. The research steps are shown in Figure 1.

2. Simulation and Measurement for the Natural Frequency of IM

Firstly, the modal analysis of the IM is carried out in ANSYS/Workbench, and the IM parameters are shown in Table 1. One end of the IM is fixed, and a fixed constraint is applied to the other end, the natural frequencies of the IM are obtained by simulation and shown in Figure 2, the 1st to 4th natural frequencies are 850 Hz, 1152 Hz, 3125 Hz and 3893 Hz, respectively.
Then, the modal experiment is carried out, the hammering method is adopted to measure the IM natural frequencies, and the vibration intensity of the IM at different frequencies is shown in Figure 3. There are four vibration peak values, the corresponding frequencies are very close to the simulation results.

3. Analysis of Harmonic Characteristics Caused by Nonlinear Factors

Figure 4 shows the topology of the IM drive system fed by the three-phase voltage source inverter (VSI), where udc is the DC-link voltage, Cdc is the DC-link capacitor, S1–S6 is the power switching tube and ia, ib, ic are the three-phase stator currents of the induction motor.
Figure 5 shows the low-order harmonic distribution of the line-to-line stator voltage uab and the stator current ias using SVPWM without deadtime and conduction voltage drop of switching devices. Except for the fundamental component, the harmonics of uab only include the high-order harmonic components around the switching frequency and its integer multiples. So only a few low-order harmonic components exist in the ias.
If the switching on delay ton and off delay toff are considered, the dead time td should be added to prevent the short circuit of the DC battery, the actual switching signals of S1 and S4 are noted as p1 and p4, which are shown in Figure 6, an obvious timing error can be seen. At present, the low-order harmonic distribution of the stator current is shown in Figure 7a. Compared to Figure 5b, the 5th and 7th harmonics increase significantly. After setting the conduction voltage drops of the switching tube and body diode, noted as uvt and uvd, respectively, the 5th and 7th harmonics further increase, which is shown in Figure 7b. The output voltage uan of phase a in a switching period Ts is shown in Figure 6, there is a nonlinear error ∆uan between uan and the ideal voltage u an * , and ∆uan is affected by the polarity of ias, which is expressed as:
Δ u an = { [ τ u dc + ( d a τ ) u vt + ( 1 d a + τ ) u vd ] ( i a > 0 ) τ u dc + ( 1 d a τ ) u vt + ( d a + τ ) u vd ( i a < 0 )
where τ = (td + tontoff)/Ts, da is the duty cycle. These low-order harmonics will increase the torque ripple. If the frequencies of these lower-order harmonics are close to the natural frequencies of the IM, the resonance will occur and the mechanical performance of the IM will deteriorate.
The sequence of the three-phase 5th harmonic is negative, the expressions of the 5th harmonic component of the three-phase stator current are defined as follows:
{ i a 5 = I 5 sin ( 5 ω t + φ 5 ) i b 5 = I 5 sin ( 5 ω t + φ 5 + 120 ) i c 5 = I 5 sin ( 5 ω t + φ 5 120 )
where I5 and φ5 are the amplitude and phase of the 5th harmonic component of the stator current, ω is the fundamental angular frequency, respectively. After the coordinate transformation based on the rotor flux orientation, the expressions of the 5th harmonic component of the stator current in d-q axis are obtained as:
{ i d 5 = I 5 cos ( 6 ω t + φ 5 ) i q 5 = I 5 sin ( 6 ω t + φ 5 )
The sequence of the three-phase 7th harmonic is positive, the expressions of the 7th harmonic component of the three-phase stator current are defined as follows:
{ i a 7 = I 7 sin ( 5 ω t + φ 7 ) i b 7 = I 7 sin ( 5 ω t + φ 7 120 ) i c 7 = I 7 sin ( 5 ω t + φ 7 + 120 )
where I7 and φ7 are the amplitude and phase of the 7th harmonic component of the stator current, respectively. The expressions of the 7th harmonic component of the stator current in d-q axis are expressed as:
{ i d 7 = I 7 cos ( 6 ω t + φ 7 ) i q 7 = I 7 sin ( 6 ω t + φ 7 )
Figure 8 shows the low-order harmonics distribution of the stator current in d-q axis, except DC component, the 6th harmonic content is obviously higher than other harmonics in id and iq. Therefore, suppression of the 6th harmonic component of id and iq is equivalent to reducing the 5th and 7th harmonic components of the three-phase stator currents.

4. Design of Current Loop Controller with 6th Harmonic Suppression

In order to reduce the 6th harmonic content in id and iq, in this paper, a 6th harmonic resonance controller GR(s) is added to the current loop controller, and the expression of GR(s) is as follows:
G R ( s ) = 2 k r ξ ω n s s 2 + 2 ξ ω n s + ω n 2
where ωn is the resonant frequency, and it is also set as the 6th harmonic angular frequency, ζ is the damping ratio, and kr is the coefficient of GR(s). Ignoring the influence of the cross-coupling in d-q axis, the control diagram of the current loop d-axis is shown in Figure 9.
Where Td is the delay time and set as 100 µs in this paper, σ is the leakage coefficient, Ls is the stator equivalent inductance, Rs is the stator resistance of the IM, respectively.

4.1. The Parameters Design of PI Controller

After zero and pole cancellation, the current loop is simplified to a type-I system, the bandwidth of the current inner loop is set as one-tenth of the switching frequency, and the parameters of the PI controller are as follows:
{ k p = 2000 π σ L s k i = 2000 π R s
The open loop transfer function Gi_ol(s) of the current loop without GR(s) can be expressed as:
G i _ ol ( s ) = 2000 π s ( T d s + 1 )
The close loop transfer function Gi_cl(s) of the current loop can be derived as:
G i _ cl ( s ) = G i _ ol ( s ) 1 + G i _ ol ( s ) = 2000 π T d s 2 + s + 2000 π
The bode plot of Gi_cl(s) is shown in Figure 10a, the bandwidth is 1 kHz, which is one-tenth of the switching frequency. The 6th harmonic distortion caused by the deadtime and conduction voltage drop in d-q axis is regarded as a disturbance ud6, the expression of the transfer function Gu_i(s) from ud6 to id can be derived as:
G u _ i ( s ) = s ( T d s + 1 ) σ L s T d s 3 + ( R s T d + σ L s ) s 2 + R s s + 2000 π
The bode plot of Gu_i(s) is shown in Figure 10b, it can be seen that the amplitude of Gu_i(s) in the range of 70 Hz to 1.8 kHz is above 0 dB line, so the PI controller is not able to suppress the low order harmonic distortion of the stator current caused by the disturbance voltage in the range of 70 Hz to 1.8 kHz.

4.2. The Parameters Design of Resonance Controller

Next, the influence of different kr on control performance will be discussed. When the fundamental frequency of the stator current is 167 Hz, the 5th and 7th harmonic frequencies are 835 Hz and 1169 Hz, respectively, which coincides with the 1st and 2nd natural frequencies of the IM. So ωn in GR(s) is set to 1002π rad/s, which is equal to six times the fundamental angular frequency, and ζ is set as 0.5. When GR(s) is introduced, the bode plot of Gi_ol(s) under different kr is shown in Figure 11. It can be seen that the addition of GR(s) has no impact on the low-frequency characteristic of Gi_ol(s), but with the increase of kr, the amplitude around 1000 Hz increases significantly, the cut-off frequency also increases as the kr, but the phase margin of Gi_ol(s) decreases, which may lead to the system instability.
To further analyze the influence of kr on stability, when kr increases from 1 to 4, the pole distribution diagram of Gi_cl(s) is shown in Figure 12. As kr increases, a pair of conjugate poles move to the right. When kr > 3.5, Gi_cl(s) appears on the right half-plane pole, which leads to system instability.
In order to theoretically prove the suppression effect of GR(s) on the 6th harmonic component in id, when kr increases from 1 to 3, the bode plot of Gu_i(s) is shown in Figure 13. Due to the addition of GR(s), the amplitude around 1000 Hz is below 0 dB line. And with the increase of kr, the suppression effect on the 6th harmonic component in id is more significant, and the band of the voltage harmonic frequency that can be suppressed by the GR(s) will also expand. Although another peak appears, the amplitude of the harmonic components of the inverter output voltage in d-q axis is very few, so it will not cause additional harmonic distortion. However, to prevent increasing the eleventh and thirteenth harmonics of stator current, the cut-off frequency of Gu_i(s) should be less than the twelfth harmonic frequency.

4.3. Analysis of the Resonance Controller under Frequency Variation

According to the above analysis presented in Section 4.2, the suppression effect can be improved with the increase of kr, however, the stability of the current loop is reduced, so kr is set to 2.5 in this paper, and the transfer function of GR(s) is shown as:
G R ( s ) = 2505 π s s 2 + 1002 π s + ( 1002 π ) 2
The bode plot of Gu_i(s) is shown in Figure 14. It can be seen that the resonance controller can suppress the stator voltage whose 6th harmonic frequency is between 0.35 kHz and 1.32 kHz. However, if the 6th harmonic frequency is between 0.065 kHz and 0.35 kHz, the corresponding fundamental frequency is between 10.8 Hz and 58.3 Hz, the resonance controller under kr =2.5 and ωn = 1002π rad/s is not able to suppress the harmonic distortion caused by the deadtime and the voltage drops of switching devices and diode.
In the case where the fundamental frequency is less than 60 Hz. Another resonance controller is designed and adopted. In consideration of the stability and the harmonic suppression effect, according to the bode plot and the pole distribution diagram, the transfer function GR(s) is rewritten as:
G R ( s ) = 600 π s s 2 + 100 π s + 10000 π 2
At this time, the bode plots of Gi_ol(s), Gi_cl(s), and Gu_i(s) are shown in Figure 15. It can be seen that the cut-off frequency of Gi_ol(s) is 1 kHz, the phase margin of Gi_ol(s) is 45°, and Gi_cl(s) has a good DC following characteristics. Meanwhile, the resonance controller shown in Formula (10) can effectively suppress the disturbance from the stator voltage whose 6th harmonic frequency is less than 0.38 kHz.
Since the two resonant controllers shown in Formulas (11) and (12) are applicable to different frequency ranges, the resonant controller needs to be selected according to the fundamental frequency. The fundamental angle frequency ωs can be calculated by the follows:
ω s = n ω r + L m R r i q L r ψ r
where Rr is the rotor resistance, ψr is the rotor flux, ωr is the speed, and n is the number of pole pairs. Lm and Lr are the mutual inductance and rotor inductance, respectively. The fundamental frequency fs = ωs/(2π), if fs > 60 Hz, the GR(s) shown as (11) is adopted, otherwise, the GR(s) shown as (12) is adopted.

5. Simulation Results

In order to verify the suppression effect of the proposed method in this paper on the low-order harmonics in the stator current, the simulation model is established in the environment of Matlab/Simulink. The IM parameters are the same as in Table 1, other parameters are set in Table 2.
In order to unify with the digital control in the experiment, the controller composed PI and resonance is implemented by a Matlab function module in the environment of Matlab/Simulink, and the sampling time is set as 100 µs. According to the difference equation, the discrete control algorithm is written in the Matlab function module, which is shown as:
u d * ( k ) = ( 3 2 ξ ω n T s ) u d * ( k 1 ) ( 1 2 ξ ω n T s + ω n 2 T s 2 ) u d * ( k 2 ) + ( k i T s + k p ) e ( k ) + 2 k r ξ ω n T s e ( k 1 ) 2 k r ξ ω n e ( k 2 )
where e is the current error. The fundamental frequency and the load are set as 167 Hz and N·m, respectively. Three different simulations are carried out under no compensation, pulse time compensation, and the 6th harmonic controller in d-q axis proposed in this paper, respectively. The simulation results under different conditions are shown in Figure 16, Figure 17 and Figure 18, respectively. The THD, 5th harmonic, and 7th harmonic of the stator current, and the torque ripple, are summarized in Table 3; it can be seen that the low order harmonics suppression effect of the 6th harmonic controller in d-q axis is optimal.
To certify the effectiveness of the resonant controller shown in Formula (12) at low frequency, the fundamental frequency is set as 30 Hz, two simulations are carried out by the controller (11) and (12), respectively. The low-order harmonics distribution of the stator current is shown in Figure 19. It is obvious that the resonant controller of (11) cannot suppress the 5th and 7th harmonic components of the stator current. As opposed to this, the 5th and 7th harmonic components are significantly reduced when the resonant controller of (12) is adopted.
Next, the impact on the dynamic performance of the system from the resonant controller is verified. Firstly, the load torque is set to 4 N·m constantly, and the command value of speed changes from 3000 rpm to 4500 rpm, the dynamic simulation results with and without the resonant controller are shown in Figure 20. Then, the command value of speed is set to 4000 rpm, the load torque increases from 2 N·m to 4 N·m, and the dynamic simulation results with and without the resonant controller are shown in Figure 21. After the resonant controller is introduced, only a slight difference occurs in the dynamic waveforms of speed and torque, so the resonant controller has no impact on the dynamic performance.

6. Experimental Results

The experimental prototype, shown in Figure 22, is established to verify the suppression effect of the proposed method on the low-order harmonics in the stator current. TMS320F28069 is used to execute the control algorithm, and three non-contact sensors MLX91205 is used to sample the stator current. The inverter bridge is composed by the MOSFET, the type is IPB042N10N, the DC-link voltage is 72 V, and is supplied by the battery pack. The parameters of the IM are the same as in Table 1, and the experimental parameters and operation are the same as in Table 2. The experimental results under no compensation, pulse time compensation, and the 6th harmonic controller in d-q axis are shown in Figure 23, Figure 24 and Figure 25, respectively. Under the influence of deadtime and conduction voltage drop, without compensation, the stator current has serious harmonic distortion, the 5th harmonic content is 3.2%, the 7th harmonic content is 2.3%, and there is a little DC component. After the pulse time compensation, the content of the 5th and 7th harmonics and THD are slightly reduced, and the power spectral density at the natural frequency drops below −10 dB/Hz. By adopting the 6th harmonic controller in d-q axis proposed in this paper, the 5th harmonic content is only 1%, the 7th harmonic content 0.6%, and the THD decreases to 4.62%. At the same time, the DC component of the current is eliminated, and the power spectral density at all frequencies is less than −20 dB/Hz. The experimental results are consistent with the simulation results, which show that the proposed resonant controller in d-q axis can reduce the low-order harmonics of the stator current.
To verify the effectiveness of the control algorithm in reducing vibration, a vibration measurement instrument is used to measure the vibration of the IM, and the measurement experiment platform is shown in Figure 26.
Figure 27 shows the power spectral density of the vibration within 1.8 kHz under no compensation, there are two peaks of vibration at the 1st and 2nd natural frequencies. Figure 28 shows the power spectral density of the vibration within 1.8 kHz under the 6th harmonic controller in d-q axis, the vibrations are significantly suppressed. The whole experimental results further prove that the 6th harmonic controller with its parameter design method in d-q axis has a good suppression effect on the 5th and 7th harmonics distortion of the stator current, which can not only reduce the torque ripple, but also weaken the vibration of the IM.

7. Conclusions

The simulation and experimental results prove that the proposed 6th harmonic controller with its parameter design method in d-q axis has good effects on reducing the 5th and 7th harmonic distortion of the stator current, weakening the low-frequency resonance of the motor, and improving the dynamic performance of the motor drive system. Finally, the following conclusions are formed:
(1)
The dead time and the conduction voltage drop of the switching tube and body diode will produce nonlinear errors in the inverter output voltage. The current PI controller based on rotor flux orientation cannot suppress the 5th and 7th harmonic distortion of the stator current caused by the nonlinear voltage;
(2)
Based on current loop stability and the harmonic suppression effect, according to the bode plot and the pole distribution diagram, the proposed 6th harmonic controller can effectively suppress the low-order harmonic distortion of the nonlinear voltage, without generating harmonic distortion of other frequencies, and does not affect the control performance of the current loop;
(3)
Since another resonant controller for the low-speed range is introduced, low-order harmonic suppression at all operating frequencies can be achieved;
(4)
Suppressing the 6th harmonic component of the stator current in d-q axis is an effective means to weaken the low-frequency resonance and vibration of the motor.

Author Contributions

Conceptualization and methodology, P.S.; simulation, Y.L.; validation and experiment, P.S. and T.L.; writing and supervision, L.W. and T.L.; review and editing, P.S.; funding acquisition, H.W. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by “the Fundamental Research Funds for the Central Universities”, Southwest Minzu University (Funder, Huazhang Wang; Funding number, 2021101).

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Du, G.; Zou, Y.; Zhang, X.; Guo, L.; Guo, N. Heuristic Energy Management Strategy of Hybrid Electric Vehicle Based on Deep Reinforcement Learning with Accelerated Gradient Optimization. IEEE Trans. Transp. Electrif. 2021, 7, 2194–2208. [Google Scholar] [CrossRef]
  2. Liang, J.; Howey, B.; Bilgin, B.; Emadi, A. Source of Acoustic Noise in a 12/16 External-Rotor Switched Reluctance Motor: Stator Tangential Vibration and Rotor Radial Vibration. IEEE Open J. Ind. Appl. 2020, 1, 63–73. [Google Scholar] [CrossRef]
  3. Miao, Y.; Liu, H.; Zhang, W.; Liu, P. Optimal Control Strategy of Storage Inductor Current and Rotor Flux for Current Source Inverter Motor Drive System. Proc. CSEE 2019, 39, 2757–2767. [Google Scholar]
  4. Song, P.; Liu, Y.; Liu, C. Research on Parameter Design and Control Method for Current Source Inverter–Fed IM Drive Systems. Machines 2022, 10, 922. [Google Scholar] [CrossRef]
  5. Jiang, Z.; Ying, D.; Jian, S. Electromagnetic Noise Characteristics of Permanent Magnet Synchronous Motor Applied in Electric Vehicle. Trans. China Electrotech. Soc. 2016, 31, 53–59. [Google Scholar]
  6. Dong, Z.; Liu, C.; Song, Z.; Liu, S. Suppression of Dual−Harmonic Components for Five−Phase Series−Winding PMSM. IEEE Trans. Transp. Electrif. 2022, 8, 121–134. [Google Scholar] [CrossRef]
  7. Lai, Y.-S.; Chang, Y.-T.; Chen, B.-Y. Novel Random-Switching PWM Technique with Constant Sampling Frequency and Constant Inductor Average Current for Digitally Controlled Converter. IEEE Trans. Ind. Electron. 2013, 60, 3126–3135. [Google Scholar] [CrossRef]
  8. Peng, H.; Heping, L.; Yiru, M.; You, X.; Guo, Q. Variable Switching Frequency Pulse Width Modulation for Induction Motors based on Current Ripple Peak Value. Trans. China Electrotech. Soc. 2020, 35, 4373–4383. [Google Scholar]
  9. Liu, P.; Xu, J.; Sun, M.; Yuan, J.; Blaabjerg, F. New Discontinuous Space Vector Modulation Strategies for Impedance-Source Inverter with Superior Thermal and Harmonic Performance. IEEE Trans. Ind. Electron. 2022, 69, 13079–13089. [Google Scholar] [CrossRef]
  10. Guo, Q.; Dong, Z.; Liu, H.; You, X. Nonlinear Characteristics Compensation of Inverter for Low-Voltage Delta-Connected Induction Motor. Energies 2020, 13, 590. [Google Scholar] [CrossRef] [Green Version]
  11. Liu, P.; Xu, J.; Yang, Y.; Wang, H.; Blaabjerg, F. Impact of Modulation Strategies on the Reliability and Harmonics of Impedance-Source Inverters. IEEE J. Emerg. Sel. Top. Power Electron. 2020, 8, 3968–3981. [Google Scholar] [CrossRef]
  12. Zhang, Z.; Xu, L. Dead-Time Compensation of Inverters Considering Snubber and Parasitic Capacitance. IEEE Trans. Power Electron. 2014, 29, 3179–3187. [Google Scholar] [CrossRef]
  13. Liu, H.; Liu, Q.; Zhang, W.; Miao, Y.; Liu, P. Random PWM Technique for Acoustic Noise and Vibration Reduction in Induction Motors used by Electric Vehicles. Trans. China Electrotech. Soc. 2019, 34, 1488–1495. [Google Scholar]
  14. Kobayashi, T.; Tajima, F.; Ito, M.; Shibukawa, S. Effects of Slot Combination on Acoustic Noise from Induction Motors. IEEE Trans. Magn. 1997, 33, 2101–2104. [Google Scholar] [CrossRef]
  15. Gao, X.; Su, D. Suppression of a Certain Vehicle Electrical Field and Magnetic Field Radiation Resonance Point. IEEE Trans. Veh. Technol. 2018, 67, 226–234. [Google Scholar] [CrossRef]
  16. Wu, Z.; Ding, K.; Yang, Z.; He, G. Analytical Prediction and Minimization of Deadtime-Related Harmonics in Permanent Magnet Synchronous Motor. IEEE Trans. Ind. Electron. 2021, 68, 7736–7746. [Google Scholar] [CrossRef]
  17. Liu, H.; Lu, Y.; Wang, H.; Miao, Y. Dead-time Compensation Modulation Strategy of Subsection Integrated in Voltage Source Inverter. Electr. Mach. Control 2018, 22, 25–32. [Google Scholar]
  18. Miao, Y.; Liu, H.; Wang, H.; Lu, Y. Average Current Feedforward Compensation Strategy of Overlap-Time Effect for Current Source Inverter. Proc. CSEE 2018, 38, 4183–4193. [Google Scholar]
  19. Liao, Y.; Zhen, S.; Liu, R.; Yao, J. Torque Ripple Suppression of Permanent Magnet Synchronous Motor by the Harmonic Injection. Proc. CSEE 2011, 21, 119–127. [Google Scholar]
  20. Yang, Y.; Zhou, K.; Wang, H.; Blaabjerg, F.; Wang, D.; Zhang, B. Frequency Adaptive Selective Harmonic Control for Grid-Connected Inverters. IEEE Trans. Power Electron. 2015, 30, 3912–3924. [Google Scholar] [CrossRef]
  21. da Silveira, D.B.; Farias, J.V.M.; Pereira, H.A.; Luiz, A.-S.A.; Stopa, M.M.; Cupertino, A.F. Mitigation of torque oscillations caused by dc injection-based resistance estimation through resonant controllers. In Proceedings of the 2021 Brazilian Power Electronics Conference (COBEP), João Pessoa, Brazil, 7–10 November 2021; pp. 1–6. [Google Scholar]
Figure 1. The chart of the research steps for this paper.
Figure 1. The chart of the research steps for this paper.
Machines 10 01206 g001
Figure 2. Simulation of the IM natural frequency. (a) 1st order modal; (b) 2nd order modal; (c) 3rd order modal; and (d) 4th order modal.
Figure 2. Simulation of the IM natural frequency. (a) 1st order modal; (b) 2nd order modal; (c) 3rd order modal; and (d) 4th order modal.
Machines 10 01206 g002
Figure 3. Experimental test of the IM natural frequency.
Figure 3. Experimental test of the IM natural frequency.
Machines 10 01206 g003
Figure 4. The topology of VSI-fed IM drive system.
Figure 4. The topology of VSI-fed IM drive system.
Machines 10 01206 g004
Figure 5. The harmonics distribution without deadtime and conduction voltage drop. (a) uab; (b) ias.
Figure 5. The harmonics distribution without deadtime and conduction voltage drop. (a) uab; (b) ias.
Machines 10 01206 g005
Figure 6. The effect diagram of the deadtime and voltage drops of the switching tube.
Figure 6. The effect diagram of the deadtime and voltage drops of the switching tube.
Machines 10 01206 g006
Figure 7. The lower order harmonics distribution of ias. (a) With deadtime; (b) with deadtime and conduction voltage drops.
Figure 7. The lower order harmonics distribution of ias. (a) With deadtime; (b) with deadtime and conduction voltage drops.
Machines 10 01206 g007
Figure 8. The harmonics distribution of the stator current in d-q axis. (a) id; (b) iq.
Figure 8. The harmonics distribution of the stator current in d-q axis. (a) id; (b) iq.
Machines 10 01206 g008
Figure 9. The control diagram of the current loop in d-axis.
Figure 9. The control diagram of the current loop in d-axis.
Machines 10 01206 g009
Figure 10. The bode plots without resonance controller. (a) Gi_cl(s); (b) Gu_i(s).
Figure 10. The bode plots without resonance controller. (a) Gi_cl(s); (b) Gu_i(s).
Machines 10 01206 g010
Figure 11. The bode plot of Gi_ol(s) with resonance controller under different kr.
Figure 11. The bode plot of Gi_ol(s) with resonance controller under different kr.
Machines 10 01206 g011
Figure 12. The pole distribution diagram of Gi_cl(s) under different kr.
Figure 12. The pole distribution diagram of Gi_cl(s) under different kr.
Machines 10 01206 g012
Figure 13. The bode plot of Gu_i(s) with resonance controller under different kr.
Figure 13. The bode plot of Gu_i(s) with resonance controller under different kr.
Machines 10 01206 g013
Figure 14. The bode plot of Gu_i(s) with resonance controller under kr = 2.5 and ωn = 1002π rad/s.
Figure 14. The bode plot of Gu_i(s) with resonance controller under kr = 2.5 and ωn = 1002π rad/s.
Machines 10 01206 g014
Figure 15. The bode plots the fundamental frequency is less than 60 Hz. (a) Gi_ol(s); (b) Gi_cl(s); and (c) Gu_i(s).
Figure 15. The bode plots the fundamental frequency is less than 60 Hz. (a) Gi_ol(s); (b) Gi_cl(s); and (c) Gu_i(s).
Machines 10 01206 g015
Figure 16. Simulation results under no compensation. (a) ias; (b) FFT; and (c) output torque.
Figure 16. Simulation results under no compensation. (a) ias; (b) FFT; and (c) output torque.
Machines 10 01206 g016
Figure 17. Simulation results under pulse time compensation. (a) ias; (b) FFT; and (c) output torque.
Figure 17. Simulation results under pulse time compensation. (a) ias; (b) FFT; and (c) output torque.
Machines 10 01206 g017
Figure 18. Simulation results under 6th harmonic controller. (a) ias; (b) FFT; and (c) output torque.
Figure 18. Simulation results under 6th harmonic controller. (a) ias; (b) FFT; and (c) output torque.
Machines 10 01206 g018
Figure 19. The FFT analysis results of the stator current under different resonant controllers. (a) GR(s) of (11); (b) GR(s) of (12).
Figure 19. The FFT analysis results of the stator current under different resonant controllers. (a) GR(s) of (11); (b) GR(s) of (12).
Machines 10 01206 g019
Figure 20. The dynamic simulation waveforms when the speed changes from 3000 rpm to 4500 rpm. (a) Without resonant controller; (b) with resonant controller.
Figure 20. The dynamic simulation waveforms when the speed changes from 3000 rpm to 4500 rpm. (a) Without resonant controller; (b) with resonant controller.
Machines 10 01206 g020
Figure 21. The dynamic simulation waveforms when the load changes from 2 N·m to 4 N·m. (a) Without resonant controller; (b) with resonant controller.
Figure 21. The dynamic simulation waveforms when the load changes from 2 N·m to 4 N·m. (a) Without resonant controller; (b) with resonant controller.
Machines 10 01206 g021
Figure 22. Experimental prototype.
Figure 22. Experimental prototype.
Machines 10 01206 g022
Figure 23. Experimental results under no compensation. (a) ias; (b) FFT; and (c) power spectral density.
Figure 23. Experimental results under no compensation. (a) ias; (b) FFT; and (c) power spectral density.
Machines 10 01206 g023
Figure 24. Experimental results under pulse time compensation. (a) ias; (b) FFT; (c) Power spectral density.
Figure 24. Experimental results under pulse time compensation. (a) ias; (b) FFT; (c) Power spectral density.
Machines 10 01206 g024
Figure 25. Experimental results under 6th harmonic controller. (a) ias; (b) FFT; and (c) power spectral density.
Figure 25. Experimental results under 6th harmonic controller. (a) ias; (b) FFT; and (c) power spectral density.
Machines 10 01206 g025
Figure 26. The vibration measurement platform.
Figure 26. The vibration measurement platform.
Machines 10 01206 g026
Figure 27. The power spectral density of vibration in IM without compensation.
Figure 27. The power spectral density of vibration in IM without compensation.
Machines 10 01206 g027
Figure 28. The power spectral density of vibration in IM 6th harmonic controller.
Figure 28. The power spectral density of vibration in IM 6th harmonic controller.
Machines 10 01206 g028
Table 1. The parameters of the IM.
Table 1. The parameters of the IM.
ParametersValueParametersValue
Rated voltage48 VRotor outer diameter109 mm
Rated power10 kWRotor inner diameter36 mm
Maximum speed6000 rpmRotor slot number42
Pole pairs2Iron core length180 mm
stator Outer diameter 188 mmStator resistance0.047 Ω
Stator inner diameter110 mmStator leakage inductance81.5 µF
stator slot number36Rotor resistance0.028 Ω
Rotor leakage inductance81.3 µFExcitation inductance2.29 mH
Table 2. Simulation parameters.
Table 2. Simulation parameters.
ParametersValueParametersValue
Switching cycle10 kHzSimulation step size0.5 μs
Sampling cycle10 kHzVoltage drop of the switching devices0.5 V
Deadtime2 μsVoltage drop of the diode0.7 V
Table 3. Comparison of the simulation results.
Table 3. Comparison of the simulation results.
Compensation MethodTHD5th Harmonic7th HarmonicTorque Ripple (N·m)
No compensation5.88%2.9%1.4%3.6 to 4.3
Pulse time compensation5.33%2.3%1.2%3.7 to 4.3
6th harmonic controller4.34%0.58%0.43%3.8 to 4.2
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Song, P.; Liu, Y.; Liu, T.; Wang, H.; Wang, L. A Novel Suppression Method for Low-Order Harmonics Causing Resonance of Induction Motor. Machines 2022, 10, 1206. https://doi.org/10.3390/machines10121206

AMA Style

Song P, Liu Y, Liu T, Wang H, Wang L. A Novel Suppression Method for Low-Order Harmonics Causing Resonance of Induction Motor. Machines. 2022; 10(12):1206. https://doi.org/10.3390/machines10121206

Chicago/Turabian Style

Song, Pengyun, Yanghui Liu, Tao Liu, Huazhang Wang, and Liwei Wang. 2022. "A Novel Suppression Method for Low-Order Harmonics Causing Resonance of Induction Motor" Machines 10, no. 12: 1206. https://doi.org/10.3390/machines10121206

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop