| Author: | Tang, Xiaomei |
| Title: | Research on control strategy for PMSM drives based on finite set model predictive control |
| Advisors: | Niu, Shuangxia (EEE) |
| Degree: | Ph.D. |
| Year: | 2025 |
| Department: | Department of Electrical and Electronic Engineering |
| Pages: | 141 pages : color illustrations |
| Language: | English |
| Abstract: | The Permanent Magnet Synchronous Motor (PMSM) has garnered extensive research and application in fields such as the defense industry and transportation due to its advantages, including high efficiency, high power density, high torque-to-current ratio, and a wide speed regulation range. Concurrently, with the advancement of powerful and fast microprocessors, Model Predictive Control (MPC) has emerged as a promising control strategy for AC motors. MPC features a straightforward control structure, excellent dynamic response, and the ability to easily account for system nonlinearities and constraints, making it particularly suitable for achieving high-performance control of PMSMs. However, the application of existing MPC algorithms to PMSM control presents a series of challenges, such as significant output ripples and elevated common-mode voltages, which can lead to decreased prediction accuracy and control performance. To address these issues, this thesis focuses on PMSMs as the research subject, employing finite set predictive control system theory and comprehensively applying modern control concepts such as hysteresis to conduct in-depth research on high-performance current control strategies for electric drives. The main contributions of this work are summarized as follows: Firstly, the mathematical models of PMSM and traditional two-level inverter are introduced, and the basic principle and solution process of finite-set MPC (FS-MPC) are described in detail. FS-MPC is combined with PMSM current control, and the entire controller replaces the current loop of traditional field-oriented control, and the cost function is designed according to the current tracking reference value. Building on this classical strategy's simulation validation, the primary challenges and limitations are discussed through theoretical derivation, encompassing dynamic and steady-state effects, output ripples, and online computation requirements. Secondly, to address the non-fixed and elevated switching frequency issues associated with FS-MPC under various operational conditions, a multi-vector switching sequence optimization is introduced into the classical control strategy. Additionally, two voltage generation methods that avoid utilizing zero vectors have been designed to suppress common-mode voltage and mitigate adverse effects such as shaft currents and electromagnetic interference. Experimental results demonstrate that this strategy can effectively accommodate the motor's varied operational states, combining the favorable steady-state performance of the multi-vector strategy with the rapid transient response of the single-vector strategy; specific pulse generation methods can effectively eliminate common-mode voltage. Furthermore, in light of the prolonged computation times and high implementation complexities associated with the three-level FS-MPC control strategy in practical hardware execution, a simplified sub-sector division has been established. A rapid optimization search method has been devised by designing a novel cost function and narrowing the traditional enumeration optimization search range. Comparative simulations with typical schemes from the literature indicate that the proposed approach is both simple to implement and capable of quickly tracking given values under dynamic conditions. Additionally, employing a pre-refined candidate vector set as the core framework, an enhanced multi-vector control strategy is proposed, which achieves neutral point potential balancing and common-mode voltage suppression. This strategy also considers optimal dwell times and the composite effects of reconstructed vectors across different sectors. Experimental results validate that this scheme can achieve satisfactory dynamic and static performance within a short execution time, while lowering the switching frequency of the three-level inverter and enhancing the operational efficiency of the drive system. Finally, a set of coherent voltage vectors (CVVs) with movable starting points is introduced to replace the basic candidates. The pulse train of the optimal CVV is generated by single-carrier modulation, and capacitor charge balancing in different sectors can be included in the zero-sequence component injection. The proposed CVV-MPC is characterized by simple implementation and satisfactory performance under low switching frequency. Comparative experiments are conducted to verify the effectiveness and superiority of the proposed method. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14097