Development of Advanced Torque Control Systems for Permanent Magnet Synchronous Machines
永磁同步電機先進轉矩控制系統的研發
Student thesis: Doctoral Thesis
Author(s)
Related Research Unit(s)
Detail(s)
Awarding Institution | |
---|---|
Supervisors/Advisors |
|
Award date | 21 Jun 2023 |
Link(s)
Permanent Link | https://scholars.cityu.edu.hk/en/theses/theses(fcb30b65-ffb5-48b0-876a-27a119402210).html |
---|---|
Other link(s) | Links |
Abstract
Permanent magnet synchronous machines (PMSMs) have been widely used and analyzed in industry and academia due to their distinct merits of high efficiency and high power density. When applying PMSMs in a practical energy conversion system, the torque output performance is one of the most critical indicators. Fast response and small ripples of torque output are usually required. To achieve this, three aspects need to be considered: torque control algorithm, harmonic currents suppression, and modulation scheme. Although the existing control and modulation methods can support stable torque output, their performance still needs improvement. Therefore, this thesis investigates high-performance torque control strategies for PMSMs considering harmonic currents suppression and optimized modulation scheme.
As for torque control, the existing most advanced method should be deadbeat-direct torque and flux control (DB-DTFC). The problem with this method is the high computational burden and machine parameter reliance. This thesis proposes a simplified DB-DTFC based on stator flux differential. This simplified method reduces the order of conventional DB-DTFC and thus lightens the computational burden. Furthermore, modified stator flux and electromagnetic torque observers are designed based on disturbance observer theory to enhance parameter robustness. In addition, a new method called deadbeat-direct complex torque control (DB-DCTC) is proposed with a newly designed parameter, reactive torque. The proposed DB-DCTC can not only control torque in a deadbeat pattern but also regulate the power factor of PMSMs by controlling reactive torque.
Harmonic currents can cause torque ripples and additional loss. Specific methods for three-phase PMSMs and multiphase PMSMs are proposed in this thesis to suppress harmonic currents. As for three-phase PMSMs, the harmonic currents and fundamental current are coupled together. Thus, a harmonic currents extraction and compensation method is designed based on reference frame transformation and specific filters. As for multiphase PMSMs, harmonic currents are located in respective harmonic spaces. This thesis proposes a deadbeat predictive harmonic current control (DPHCC) method in the harmonic spaces to suppress harmonic currents. The DPHCC method can directly calculate the harmonic control voltage based on the sampled harmonic currents and the machine model. Thus, the steady-state and transient-state performance of harmonic current control can be improved. Further, a disturbance observer is designed according to the discrete-time sliding mode theory to avoid the influence of inaccurate machine parameters and other unmodeled disturbances.
Modulation schemes are necessary to synthesize the control voltage to drive PMSMs. In multiphase PMSM control systems or multiple-PMSMs control systems, since there are several control voltage spaces, synthesizing control voltage properly in each space is crucial. This thesis adopts nonlinear optimization theory to design advanced modulation schemes for multiphase PMSM control systems and double-PMSM control systems. In multiphase PMSM control systems, the voltage space corresponding to torque output is the highest priority, and the control voltage in this space is firstly synthesized. Then, optimal control voltages are found in the other spaces based on the optimization algorithm. In double-PMSM control systems, the priority of the PMSMs is determined by their required common leg duty cycle. Then, the duty cycle of the high-priority PMSM is guaranteed firstly, and then the duty cycle of the low-priority PMSM is calculated with the optimization algorithm.
In addition, all the proposed control and modulation methods in this thesis have been verified with experimental results based on several specific PMSM control systems. These methods improve the torque output performance of PMSM and can be extended to the industry in the future.
As for torque control, the existing most advanced method should be deadbeat-direct torque and flux control (DB-DTFC). The problem with this method is the high computational burden and machine parameter reliance. This thesis proposes a simplified DB-DTFC based on stator flux differential. This simplified method reduces the order of conventional DB-DTFC and thus lightens the computational burden. Furthermore, modified stator flux and electromagnetic torque observers are designed based on disturbance observer theory to enhance parameter robustness. In addition, a new method called deadbeat-direct complex torque control (DB-DCTC) is proposed with a newly designed parameter, reactive torque. The proposed DB-DCTC can not only control torque in a deadbeat pattern but also regulate the power factor of PMSMs by controlling reactive torque.
Harmonic currents can cause torque ripples and additional loss. Specific methods for three-phase PMSMs and multiphase PMSMs are proposed in this thesis to suppress harmonic currents. As for three-phase PMSMs, the harmonic currents and fundamental current are coupled together. Thus, a harmonic currents extraction and compensation method is designed based on reference frame transformation and specific filters. As for multiphase PMSMs, harmonic currents are located in respective harmonic spaces. This thesis proposes a deadbeat predictive harmonic current control (DPHCC) method in the harmonic spaces to suppress harmonic currents. The DPHCC method can directly calculate the harmonic control voltage based on the sampled harmonic currents and the machine model. Thus, the steady-state and transient-state performance of harmonic current control can be improved. Further, a disturbance observer is designed according to the discrete-time sliding mode theory to avoid the influence of inaccurate machine parameters and other unmodeled disturbances.
Modulation schemes are necessary to synthesize the control voltage to drive PMSMs. In multiphase PMSM control systems or multiple-PMSMs control systems, since there are several control voltage spaces, synthesizing control voltage properly in each space is crucial. This thesis adopts nonlinear optimization theory to design advanced modulation schemes for multiphase PMSM control systems and double-PMSM control systems. In multiphase PMSM control systems, the voltage space corresponding to torque output is the highest priority, and the control voltage in this space is firstly synthesized. Then, optimal control voltages are found in the other spaces based on the optimization algorithm. In double-PMSM control systems, the priority of the PMSMs is determined by their required common leg duty cycle. Then, the duty cycle of the high-priority PMSM is guaranteed firstly, and then the duty cycle of the low-priority PMSM is calculated with the optimization algorithm.
In addition, all the proposed control and modulation methods in this thesis have been verified with experimental results based on several specific PMSM control systems. These methods improve the torque output performance of PMSM and can be extended to the industry in the future.