Search

With the widespread integration of machine drives in aerospace and electric vehicles, high reliability has become a paramount concern. Fault diagnosis and tolerant control play crucial roles in enhancing the dependability of machine drives. However, prevailing research on fault detection typically concentrates on a single fault. In fact, various faults might share similar fault characteristics. This limitation could result in misdiagnosis and incorrect fault-tolerant of machine drives. To address this issue, this article delves into the analysis of several common electrical faults and proposes a detection method based on high frequency (HF) and low frequency (LF) signals for identifying stator winding inter-turn fault, stator winding high resistance fault, current sensor bias fault, current sensor gain fault which may exhibit similar fault characteristics. By scrutinizing the HF and LF voltage residuals, four fault indicators at different frequencies were designed, enabling reliable fault diagnosis and differentiation. The effectiveness of the proposed method is validated through extensive experiments on a permanent magnet machine.

In this article, an online attribute matching based few-sample data-driven diagnosis method for electrical faults in permanent magnet synchronous motor drive is proposed to improve the diagnosis precision with fewer training data and lower computational complexity. By incorporating the motor model knowledge in the feature extraction process, effective and robust input features can be extracted without consuming large computing resources. Instead of directly predicting the fault mode in the original huge solution space, an integrated attribute learner is specifically designed to predict the fault attributes in the simplified solution space to reduce both network size and required training data. Finally, based on the look-up table for the fault attribute vector, an attribute matching strategy is proposed to determine the specific fault mode with the consideration of unknown faults. Comprehensive experiments verify that the proposed method can identify each fault mode efficiently with an average diagnostic accuracy of 99.12% and the computational time of 0.1 ms, which outperforms the existing data-driven diagnosis methods for motor drives.

The accuracy of mathematical model of interior permanent magnet synchronous motors (IPMSMs) is of great importance for high-performance position sensorless control. However, the eddy-current reaction has not been modeled accurately in existing position sensorless control, in particular for the high-frequency (HF) voltage signal injection based position sensorless control. In this article, a novel HF signal injection based position sensorless control method is proposed for IPMSMs under the low-speed region. The key is to design a new HF position observation model of IPMSM with the vector magnetic circuit. By incorporating the magductance in magnetic circuit, the eddy-current reaction can be taken into account effectively in the proposed HF position observation model for IPMSMs. On this basis, the perturbation items due to eddy-current reaction in HF position observer are revealed and compensated to improve the accuracy of position estimation. Finally, the proposed HF signal injection based position sensorless control strategy is verified by experiments on an IPMSM drive platform.

With the increasing application of permanent magnet (PM) machines, the drive reliability should be guaranteed for safety-critical applications. As a severe stator winding short-circuit fault, turn fault causes excessive fault current and affects the machine's stable operation. Thus, this article proposes a new turn fault detection strategy based on high-frequency (HF) neutral-to-ground voltages (NTGVs) with HF injection. The electrical signal response to the turn fault is analyzed by analytical model and the expressions of the estimated three-phase NTGVs are derived by considering the HF signals only. The estimated NTGVs are identical under healthy condition; however, this is not the case under turn fault. To eliminate the influence of speed and load, HF voltages are injected and the HF NTGVs are processed to construct the fault indicator. The calculation of NTGVs only requires phase current and drive signals from the control system, eliminating the need for additional voltage sensors and hardware circuits. The effectiveness and robustness of the proposed detection strategy are demonstrated by extensive experiments via a PM machine drive.

Aiming at mitigating the torque ripple at low-carrier-ratio operating conditions, an optimal pulsewidth modulation (PWM) scheme is proposed for the dual three-phase permanent-magnet synchronous motor (DTP-PMSM). The current trajectory induced by inverter harmonic voltages on the torque subspace is derived and the model of torque ripple is established. Then, the switching angles are optimized offline to mitigate the torque ripple directly while limiting the current harmonics simultaneously. It is revealed that the torque ripple is influenced by torque angle in addition to the voltage harmonics. Moreover, the differences between the optimal switching angles for torque ripple and those for current harmonics in DTP-PMSM are analyzed. To achieve the closed-loop control performance, the proposed optimal PWM strategy is applied to the model predictive pulse pattern control (MP3C) scheme. The experimental results are given to verify the validity of the proposed optimal PWM-based MP3C scheme.

With the ever-increasing application of permanent magnet synchronous machine (PMSM) drives, the requirements for safety and reliability are raised. This article presents a detailed analysis of the open-circuit fault behavior of the inverter power device by comparing the reference voltage and the actual output voltage under both healthy and faulty conditions. The reference voltage is predicted by real-time switching state of inverter power device while the actual output voltage is estimated by a healthy machine model. It is found that the relative β-axis residual voltage of the fault phase remains unchanged during open-circuit fault, however, it is not the case for the healthy phases. This unique fault behavior is exploited for open-circuit fault detection in the PMSM drives. The voltage residuals are used to construct fault indicator and their polarities are employed for fault location identification. Average detection time is 5% electric cycle which is much faster than the existing method. Finally, the effectiveness and robustness of the proposed technique are verified under various load and speed conditions via a triple three-phase PMSM drive.