An effective approach based on nonlinear spectrum and improved convolution neural network for analog circuit fault diagnosis.

In this work, an effective approach based on a nonlinear output frequency response function (NOFRF) and improved convolution neural network is proposed for analog circuit fault diagnosis. First, the NOFRF spectra, rather than the output of the system, are adopted as the fault information of the analog circuit. Furthermore, to further improve the accuracy and efficiency of analog circuit fault diagnosis, the batch normalization layer and the convolutional block attention module (CBAM) are introduced into the convolution neural network (CNN) to propose a CBAM-CNN, which can automatically extract the fault features from NOFRF spectra, to realize the accurate diagnosis of the analog circuit. The fault diagnosis experiments are carried out on the simulated circuit of Sallen-Key. The results demonstrate that the proposed method can not only improve the accuracy of analog circuit fault diagnosis, but also has strong anti-noise ability.

Fault mechanism analysis and diagnosis for closed-loop drive system of industrial robot based on nonlinear spectrum.

To solve the problem of nonlinear characteristics neglecting and fault mechanism analysis lacking in fault diagnosis research, a new method of fault mechanism analysis and diagnosis based on nonlinear spectrum is proposed. Firstly, based on the Permanent Magnet Synchronous Motor (PMSM) model of robot, the first 4-order spectrums based on nonlinear output frequency response function (NOFRF) in different states are obtained by batch calculation method. Secondly, the high-frequency spectrum distribution rule of NOFRF spectrum in different states are analyzed. Finally, in the closed-loop simulation environment of robot, the identification method based on data-driven is adopted for NOFRF spectrum calculation to verify power loss fault of PMSM. Meanwhile, the fault diagnosis experiment is also carried out. The experimental results indicate that the key characteristics distribution rule of NOFRF spectrums in the real environment is consistent with the theoretical analysis results, and compared with the traditional fault feature extraction methods by output signal, the diagnosis with fault feature of NOFRF spectrum for industrial robot closed-loop drive system has the highest accuracy, which verifies the validity of NOFRF spectrum as the fault feature.

The optimized algorithm based on machine learning for inverse kinematics of two painting robots with non-spherical wrist.

This paper studies the inverse kinematics of two non-spherical wrist configurations of painting robot. The simplest analytical solution of orthogonal wrist configuration is deduced in this paper for the first time. For the oblique wrist configuration, there is no analytical solution for the configuration. So it is necessary to solve by general method, which cannot achieve high precision and high speed as analytic solution. Two general methods are optimized in this paper. Firstly, the elimination method is optimized to reduce the solving speed to 20% of the original one, and the completeness of the method is supplemented. Based on the Gauss damped least squares method, a new optimization method is proposed to improve the solving speed. The enhanced step length coefficient is introduced to conduct studies with the machine learning correlation method. It has been proved that, on the basis of ensuring the stability of motion, the number of iterations can be effectively reduced and the average number of iterations can be less than 5 times, which can effectively improve the speed of solution. In the simulation and experimental environment, it is verified.

A novel method of combining generalized frequency response function and convolutional neural network for complex system fault diagnosis.

To solve the problem of low accuracy in traditional fault diagnosis methods, a novel method of combining generalized frequency response function(GFRF) and convolutional neural network(CNN) is proposed. In order to accurately characterize system state information, this paper proposed a variable step size least mean square (VSSLMS) adaptive algorithm to calculate the second-order GFRF spectrum values under normal and fault states; In order to improve the ability of fault feature extraction, a convolution neural network (CNN) with gradient descent learning rate and alternate convolution layer and pooling layer is designed to extract the fault features from GFRF spectrum. In the proposed method, the second-order GFRF spectrum of each state of Permanent Magnet Synchronous Motor (PMSM) is obtained by VSSLMS; Then, the two-dimension GFRF spectrum, which is regarded as the gray value of the image,will be further transformed into image. Finally, the CNN is trained with learning rate by gradient descent way to realize the fault diagnosis of PMSM. Experimental results indicate that the accuracy of proposed method is 98.75%, which verifies the reliability of the proposed method in application of PMSM fault diagnosis.