The Permanent Magnet Synchronous Motor (PMSM) is the power source maintaining the stable and efficient operation of various pieces of equipment; hence, its reliability is crucial to the safety of public equipment. Convolutional Neural Network (CNN) models face challenges in extracting features from PMSM current data. A new Discrete Wavelet Transform Convolutional Neural Networks (DW-CNN) feature with fusion weight updating Long Short-Term Memory (LSTM) anomaly detection is proposed in this paper. This approach combines Discrete Wavelet Transform (DWT) with high and low-frequency separation processing and LSTM. The anomaly detection method adopts DWT and CNN by separating high and low-frequency processing. Moreover, this method combines the hybrid attention mechanism to extract the multi-current signal features and detects anomalies based on weight updating the LSTM network. Experiments on the motor bearing real fault dataset and the PMSM stator fault dataset prove the method’s strong capability in fusing current features and detecting anomalies.

A novel modeling method was proposed to provide an improved representation of the actual microstructure of TiB/Ti-6Al-4V discontinuously-reinforced titanium matrix composite (DRTMC). Based on the Thiessen polygon structure, the representative volume element (RVE) containing the complex microstructures of the DRTMC was first generated. Thereafter, by using multiple user-defined subroutines in the commercial finite element software ABAQUS, the application of asymmetric mesh periodic boundary conditions on the RVE was realized, and the equivalent elastic modulus of the DRTMC was determined according to the homogenization method. Through error analyses on the experimental and calculated results regarding the equivalent elastic parameters of the DRTMC, the rationality of generating the DRTMC finite element model by using the present method was validated. Finally, simulations based on four types of network-like models revealed that the present simplifications to the particle shape of the reinforcement phase had less of an influence on the overall composite strength. Moreover, the present study demonstrates that the DRTMC enhancement is mainly attributed to the matrix strengthening, rather than the load-transferring mechanism. The strengthening influences of the distribution forms of the reinforcement phases, including their distribution density and orientation, were studied further. It was found that both the higher distribution density and limited distribution orientation of the particles would increase the probability of overlapping and merging between particles, and; therefore, higher strength could be yielded when the volume fraction of the reinforcement phase reached a certain threshold. Owing to the versatility of the developed methods and programs, this work can provide a useful reference for the characterization of the mechanical properties of other composites types.

In this paper, the vacancy formation at the interface between different grains (Ag₃Sn and βSn) induced by electromigration was investigated from the perspective of atom diffusion. To explain the micro-mechanism of void formation near the interface, the diffusion coefficient was specifically studied here via molecular dynamics (MD) simulation. By comparing the atom diffusion rates of atoms in βSn and Ag₃Sn, a significant difference could be observed when the temperature is up to 400 K. The Sn atoms in βSn have a higher diffusion coefficient (8E − 9 cm²/s) than atoms in Ag₃Sn (4E − 9 cm²/s), which indicated that the void would be prone to appear in βSn near the interface. Moreover, the effect of grain size and pressure on atom diffusivity was studied. Results show that the atom diffusivity depends heavily on the grain size of Ag₃Sn. When the thickness of Ag₃Sn is increased from 4 to 12 nm, this difference is significant when the temperature is only 375 K. On the other hand, the atom diffusion character of Ag₃Sn and βSn changes substantially under constant pressure. The difference of the atom diffusion rate would be inhibited by pressure perpendicular to the interface, which indicated voids have less possibility to appear herein.

This paper proposes a new prediction method for electromigration (EM) induced void generation of solder bumps in a wafer level chip scale package. The methodology is developed based on discretized weighted residual method in a user-defined finite element analysis framework to solve the local ME governing equation with the variable of atomic concentration. The local solution of atomic concentration is incorporated in the multiphysics environment for electrical, thermal and stress in both sub-model and global model. The new method takes the advantage of solving the variable of atomic density, it avoids directly solving the divergences of the atomic flux, which includes the atomic density gradient items and is very hard and challenging to get the solution by traditional method. Comparison of the atomic density distributions with and without considering the atomic density gradient for representive nodes is investigated. The simulation results for voids and time to failure (TTF) are discussed and correlated with previous test results. Finally, the analysis of the impact of under ball metallurgy and solder bump geometry on the void generation and TTF is presented.

In recent decades, cuffless blood pressure monitoring technology has been a point of research in the field of health monitoring and public media. Based on the web of science database, this paper evaluated the publications in the field from 1990 to 2020 using bibliometric analysis, described the developments in recent years, and presented future research prospects in the field. Through the comparative analysis of keywords, citations, H-index, journals, research institutions, national authors and reviews, this paper identified research hotspots and future research trends in the field of cuffless blood pressure monitoring. From the results of the bibliometric analysis, innovative methods such as machine learning technologies related to pulse transmit time and pulse wave analysis have been widely applied in blood pressure monitoring. The 2091 articles related to cuffless blood pressure monitoring technology were published in 1131 journals. In the future, improving the accuracy of monitoring to meet the international medical blood pressure standards, and achieving portability and miniaturization will remain the development goals of cuffless blood pressure measurement technology. The application of flexible electronics and machine learning strategy in the field will be two major development directions to guide the practical applications of cuffless blood pressure monitoring technology.

This paper presents a numerical study of electronic migration in solder joint and interconnects metal lines of IC packages. The three dimensional finite element models for interconnect structure on silicon die and a flip chip solder joint structure are developed. These models have been tested through numerical experiment to obtain the electrical, thermal and structural parameters with the migration failure under the loading of high direct current density. An improved advanced algorithm of void formation and propagation in interconnect/solder joints has been developed with consideration of electro-migration, thermo-migration and stress-migration. This improved algorithm is based on direct couple analysis that includes electrical, thermal and structural fields. Then the time to failure life predictions due to mass migration have been simulated based on the voids, current density, temperature and mechanical stress distribution. The simulation results are found to be in good agreement with the experiment data presented in previous work