In most industrial applications, it is difficult to obtain complete demagnetization fault signals of all conditions with labels for permanent magnet synchronous motor (PMSM), and motors are not allowed to be disassembled, so non-contact diagnostic methods are essential. A non-contact fault diagnosis method using magnetic leakage signal based on wavelet scattering convolution network (WSCN) and semi-supervised deep rule-based (SSDRB) classifier is proposed. Through magnetic equivalent circuit model analysis, the magnetic leakage signal on motor surface is selected as fault signal. To avoid complex signal processing, the symmetrized dot pattern method is introduced to convert fault signals into two-dimensional images. Then, WSCN is applied to extract features from images, and SSDRB classifier is adopted to diagnose demagnetization fault. Finally, faulty motor prototypes are manufactured for experiment. By comparing with other methods, the superiority and effectiveness of the proposed method using a small number of labeled samples under different conditions are verified.

This article proposes a robust deadbeat predictive current control (DPCC) method for the interior permanent magnet synchronous motor (IPMSM) based on a full parameter identification method. The DPCC method is innovatively combined with the high frequency currents injection method, which can observe inductances, flux linkage, and resistance values of the IPMSM online. After the parameter identification results are obtained, the initial crude parameters are replaced by the identification values in the DPCC method to enhance the robustness of the IPMSM control system. Moreover, a novel online accuracy verification method for parameter identification results is proposed to judge the error estimation level. The proposed method extracts the high bandwidth and the high parameter sensitivity characteristic of the DPCC method. The online identification and verification of IPMSM parameters are proceeded subsequently while achieving robust control of the IPMSM with the DPCC method. Finally, the proposed method is experimentally demonstrated with a 1.5 kW motor. The experimental results show that the IPMSM with the proposed method can still work well even in the situation of severe initial parameters mismatch.

To reduce the convergence time and improve the anti-disturbance capability of the permanent magnet synchronous motor (PMSM) drivers, a fast sliding mode speed controller based on a new compound reaching law (NCRL) and disturbance observer compensation technique is proposed. Firstly, to solve the conflict between fast convergence rate and chattering reduction in the existing solutions, the NCRL is proposed by introducing the exponential function and the power term piecewise function. The convergence of NCRL is proved and the reaching time is certified to be independent of the system’s initial state. The proposed NCRL can effectively reduce the reaching time, and smooth the output signal while maintaining the high tracking performance of the controller under different conditions. However, external load disturbances may degrade the performance of the proposed NCRL-based speed controller. The proposed controller is then combined with an improved sliding mode observer (ISMO), named the NCRL+ISMO method, to improve the estimation accuracy of the lumped disturbance and the robustness of the system. The stability of the proposed method is analyzed by the Lyapunov function. Experimental results show that the proposed method can reduce speed fluctuation and convergence time, and enhance anti-disturbance capability.

Barely visible impact damage (BVID) is inevitable during either fabrication or lifetime of a carbon fiber reinforced plastic (CFRP) component. These flaws are usually difficult to be detected from the surface by visual inspection or machine vision based on a charge-coupled device or CMOS. In order to solve the problems of low efficiency, low sensitivity, and small detection area of the existing nondestructive testing (NDT) for BVID in CFRP, this paper proposes for the first time the integrated nonlinear ultrasound (NU) and vibrothermography (VT) NDT based on the shared excitation sources. The experimental system was built after introducing the principle of shared excitation based NU&VT NDT. The CFRP plates with 5, 15, and 25 J visible impact damage (VID) as well as 12 and 16 J BVID were tested using the integrated NU&VT. Experimental studies after signal processing have shown that all VID and BVID could be detected by the integrated NU&VT NDT, and the defection capability has a significant improvement after fast Fourier transform. The proposed method could provide a visualized and effective means for quality control and inspection of large-scaled and complex shape key components in manufacturing process and in service.

In intelligent fault diagnosis, transfer learning can reduce the requirement of sufficient labeled data and the same data distribution. However, for the diagnosis of a new machine, there are still some limitations, such as low accuracy or the demand for some labeled data with fault information in the new machine. In this article, we propose a stack autoencoder transfer learning algorithm based on the class separation and domain fusion (SAE-CSDF) to solve these problems. According to the characteristics of bearing faults, the proposed weighted domain fusion strategy can ensure the direction and balance in the transfer process. The proposed class separation degree can improve the accuracy of the target domain indirectly by extending the differences between the classes in the source domain. The effectiveness of the SAE-CSDF is verified via the mutual transfer of two public datasets and one laboratory dataset. The results show that the accuracy of the algorithm can reach 97% in the transfer between different machines, even if there is no labeled fault data in the new machine.

Traditional fault diagnosis methods are easy to be affected by different working conditions. This article proposed a motor fault diagnosis method based on visual knowledge, to reduce the impact of changes in working conditions and improve the feature extraction ability. The mapping relationship between actual faults and image intuitive features by symmetrized dot pattern and scale-invariant feature transform is established in this article. The fault state is obtained by statistics of the matching point with the dictionary templates generated from signals of normal and unnormal motors. Compared with other machine learning algorithms, this method does not need too much data training and learning. The efficiency of this method is validated by experiments, and the data image processing technology has great industrial application value in the field of motor fault detection or monitoring in the age of intelligence.