Motor Fault Diagnosis Based on Convolutional Block Attention Module-Xception Lightweight Neural Network.

Electric motors play a crucial role in self-driving vehicles. Therefore, fault diagnosis in motors is important for ensuring the safety and reliability of vehicles. In order to improve fault detection performance, this paper proposes a motor fault diagnosis method based on vibration signals. Firstly, the vibration signals of each operating state of the motor at different frequencies are measured with vibration sensors. Secondly, the characteristic of Gram image coding is used to realize the coding of time domain information, and the one-dimensional vibration signals are transformed into grayscale diagrams to highlight their features. Finally, the lightweight neural network Xception is chosen as the main tool, and the attention mechanism Convolutional Block Attention Module (CBAM) is introduced into the model to enforce the importance of the characteristic information of the motor faults and realize their accurate identification. Xception is a type of convolutional neural network; its lightweight design maintains excellent performance while significantly reducing the model's order of magnitude. Without affecting the computational complexity and accuracy of the network, the CBAM attention mechanism is added, and Gram's corner field is combined with the improved lightweight neural network. The experimental results show that this model achieves a better recognition effect and faster iteration speed compared with the traditional Convolutional Neural Network (CNN), ResNet, and Xception networks.

Effects of synergistic action on rheological and thermal properties of potato starch complexes co-gelatinized with caffeic acid and squash polysaccharides extracted with water and subcritical water.

In order to broaden the application range of squash polysaccharide (WESP/SWESP) and caffeic acid (CAA) and improve the quality of potato starch (PS) products, the effects of WESP/SWESP and CAA on the gelatinization, rheology, thermodynamics, microstructure and in vitro digestion of PS were investigated. Meanwhile, the synergistic effect of WESP/SWESP and CAA on PS was further analyzed. Differently, due to WESP and SWESP had different monosaccharide composition and structure, they had different effects on the system. Pasting properties results showed that the presence of WESP/SWESP and CAA significantly reduced the peak viscosity, trough viscosity, breakdown viscosity and final viscosity of PS, especially under the combined action. In rheological tests, all sample gels belonged to the pseudoplastic fluids and weak gel system (tan δ < 1). Besides, thermodynamic properties revealed that WESP/SWESP and CAA synergistic effect had better retrogradation delay effect. In the ternary system, WESP/SWESP, CAA and PS can form a new network structure and improve the stability of the gel system. In addition, the results of infrared spectroscopy, Raman spectroscopy, x-ray diffraction and scanning electron microscopy exhibited that the ternary system can promote the accumulation and winding of the spiral structure of PS chain, and make the structure of PS gel network more orderly and stable. Furthermore, compared with PS gel, the ternary system had lower RDS and higher SDS and RS content, suggesting that the addition of WESP/SWESP and CAA at the same time was more conducive to reducing the hydrolysis rate of PS. This work revealed the interaction between WESP/SWESP, CAA and PS, which improved the physicochemical and digestive properties of PS. It will provide a theoretical basis for improving the quality of potato starch-related products and developing functional foods.

Research on Three-Phase Asynchronous Motor Fault Diagnosis Based on Multiscale Weibull Dispersion Entropy.

Three-phase asynchronous motors have a wide range of applications in the machinery industry and fault diagnosis aids in the healthy operation of a motor. In order to improve the accuracy and generalization of fault diagnosis in three-phase asynchronous motors, this paper proposes a three-phase asynchronous motor fault diagnosis method based on the combination of multiscale Weibull dispersive entropy (WB-MDE) and particle swarm optimization-support vector machine (PSO-SVM). Firstly, the Weibull distribution (WB) is used to linearize and smooth the vibration signals to obtain sharper information about the motor state. Secondly, the quantitative features of the regularity and orderliness of a given sequence are extracted using multiscale dispersion entropy (MDE). Then, a support vector machine (SVM) is used to construct a classifier, the parameters are optimized via the particle swarm optimization (PSO) algorithm, and the extracted feature vectors are fed into the optimized SVM model for classification and recognition. Finally, the accuracy and generalization of the model proposed in this paper are tested by adding raw data with Gaussian white noise with different signal-to-noise ratios and the CHIST-ERA SOON public dataset. This paper builds a three-phase asynchronous motor vibration signal experimental platform, through a piezoelectric acceleration sensor to discern the four states of the motor data, to verify the effectiveness of the proposed method. The accuracy of the collected data using the WB-MDE method proposed in this paper for feature extraction and the extracted features using the optimization of the PSO-SVM method for fault classification and identification is 100%. Additionally, the proposed model is tested for noise resistance and generalization. Finally, the superiority of the present method is verified through experiments as well as noise immunity and generalization tests.

Design and topology optimization of air conditioning suspension bracket for metro.

During the operation of subway vehicles, the vibration of air conditioning units is mainly transmitted to the vehicle body through the suspension support, which seriously affects the stability and comfort of the vehicle during operation. Therefore, the design and optimization of the suspension support of air conditioning units has become a hot topic in the research of the dynamic characteristics of subway vehicles. In this paper, the rigid and flexible coupling dynamic model of metro is firstly calculated to simulate the stress of the suspension point of air conditioning of the vehicle body when the vehicle is running. The initial structure design of the suspension support is carried out, and the stress of the air conditioning suspension point is taken as the load input to analyze the stiffness and strength of the initial structure of the suspension support. Then, the fatigue life is taken as the topology constraint, and the variable density method (SIMP) is used to optimize the topology of the suspension bracket. Finally, the optimized suspension support is validated. The results show that after topological optimization, the maximum displacement and maximum stress of the suspension support under vertical, horizontal, and vertical loads are reduced by 80%, 93%, and 99%, respectively, compared with the original structure model, and the maximum stress under vertical loads is reduced by 50%.

A de-illumination scheme for face recognition based on fast decomposition and detail feature fusion.

Almost all the face recognition algorithms are unsatisfied due to illumination variation. Feature with high frequency represents the face intrinsic structure according to the common assumption that illumination varies slowly and the face intrinsic feature varies rapidly. In this paper, we will propose an adaptive scheme based on FBEEMD and detail feature fusion. FBEEMD is a fast version of BEEMD without time-consuming surface interpolation and iteration computation. It can decompose an image into sub-images with high frequency matching detail feature and sub-images with low frequency corresponding to contour feature. However, it is difficult to determine by quantitative analysis that which sub-images with high frequency can be used for reconstructing an illumination-invariant face. Thus, two measurements are proposed to calculate weights for quantifying the detail feature. With this fusion technique, one can reconstruct a more illumination-neutral facial image to improve face recognition rate. Verification experiments using classical recognition algorithms are tested with Yale B, PIE and FERET databases. The encouraging results show that the proposed scheme is very effective when dealing with face images under variable lighting condition.