應用於表面式永磁同步馬達故障情境之功率硬體迴路模擬

Abstract

本論文主旨在於使用功率硬體迴路(Power Hardware-in-the-Loop, PHIL)技術，實現對於表面式永磁同步馬達(Surface Permanent Magnet Synchronous Motor, SPMSM)在不同故障情境下的實時模擬。論文主要分為兩個部份，第一部分是針對SPMSM常見的故障情境，包含電阻不平衡(Resistance Unbalance, R-Unbalance)、欠相故障(Open-Phase Fault, OPF)以及匝間短路故障(Inter-Turn Short-Circuit Fault, ISCF)，進行模型的建立以及簡化。

第二部分則是當待測馬達驅動器(Motor driver Under Test, MUT)使用磁場導向控制(Field-Oriented Control, FOC)時， PHIL會經由故障情境之馬達模型產生具有諧波成分的參考電流。由於此參考電流無法只透過比例積分(Proportional-Integral; PI)控制器達成無穩態誤差，因此本文提出了具耦合項之比例積分諧振(Coupling-Proportional-Integral-Resonant, CPIR)控制器，以改善PHIL電流控制器的效能。

最後，這些論述的成果將透過MATLAB & Simulink模擬進行初步驗證，並同時在硬體實驗上取得相對應的成果，以證明實作及模擬皆與理論相符。本論文從實測數據驗證了PHIL在使用CPIR控制後，直交軸上兩倍頻電流的誤差在R-Unbalance情境下從17.64%降到0.43%，在OPF情境下從17.21%降到0.39%，在ISCF情境下從15.70%降到0.47%。另外，實測結果的三相電流及數值模擬的三相電流之均方根誤差(Root-Mean-Square Error, RMSE)，在使用了CPIR控制後，R-Unbalance情境下從18.31%降到14.36%，OPF情境下從17.21%降到7.94%，ISCF情境下從20.29%降到18.34%。此外，為了公正評估PHIL模擬實際馬達的能力，本論文提出了使用總諧波失真再加上雜訊(Total Harmonic Distortion plus Noise,THD+N)來進行計算的相似指數(Simularity Index, SI)。透過SI可比較實測結果的三相電流以及數值模擬的三相電流的相似程度。在使用了CPIR控制後，R-Unbalance情境下SI從86.42%升到97.33%，OPF情境下SI從55.64%升到84.31%，ISCF情境下SI從61.43%升到64.07%。透過上述三種比較數據可驗證使用CPIR控制的PHIL在故障情境下的實時模擬效能。

The objective of this thesis is to realize the emulation of Surface Permanent Magnet Synchronous Motor (SPMSM) under different fault scenarios using Power Hardware-in-the-Loop (PHIL) technology. This thesis is divided into two main parts. The first part involves the development and simplification of models for common SPMSM fault scenarios, including Resistance Unbalance(R-Unbalance), Open-Phase Fault (OPF), and Inter-Turn Short-Circuit Fault (ISCF).

The second part focuses on the emulation of faulty motors using a Motor Driver Under Test (MUT) with Field-Oriented Control (FOC). The motor model considering fault conditions generates the reference currents with harmonic components, where the conventional Proportional-Integral (PI) control cannot achieve zero steady-state error. To improve the performance of the PHIL current control, this thesis proposes a new method called Coupling-Proportional-Integral-Resonant (CPIR) control.

Finally, the derived PMSM fault models and the proposed CPIR control are preliminarily verified using MATLAB & Simulink simulations. Also, hardware experiments are conducted to validate that both simulation and experiment results are consistent. Furthermore, this thesis provides a comprehensive analysis of the error at second order harmonic of the controller. It is evident that the implementation of CPIR control leads to a significant reduction in the error at second order harmonic in various fault scenarios. Specifically, in the case of R-Unbalance scenario, the error drops from 17.64% to 0.43%, in the OPF scenario, it decreases from 17.21% to 0.39%, and in the ISCF scenario, it diminishes from 15.70% to 0.47%. Additionally, a comparison is conducted between the root-mean-square errors (RMSE) of the three-phase current measured from hardware experiments and the three-phase current obtained from numerical simulation. The results demonstrate notable improvements with the application of CPIR control. In the R-Unbalance scenario, RMSE decreases from 18.31% to 14.36%, in the OPF scenario, it drops from 17.21% to 7.94%, and in the ISCF scenario, it goes from 20.29% to 18.34%. To further assess the effectiveness of the PHIL-based motor test-bench, this thesis proposes the use of the Similarity Index (SI), calculated based on the Total Harmonic Distortion plus Noise (THD+N), to quantify the similarity between the three-phase currents of experimental and numerical simulation. The results reveal substantial enhancements in similarity with the adoption of CPIR control. In the R-Unbalance scenario, the SI increases from 86.42% to 97.33%, in the OPF scenario, the SI rises from 55.64% to 84.31%, and in the ISCF scenario, it elevates from 61.43% to 64.07%. These three hardware result comparisons confirm the effectiveness of PHIL under fault scenarios.