Search

In this article, we propose an interactive visualization method for motor design. We have proposed a visualization system for education using augmented reality (AR) and virtual reality (VR). As a further development of this, we are working on motor design as an industrial application. There are two issues in achieving high-speed interactivity. One is a fast mesh modification method in finite element analysis, and the other is a fast solving method. We propose a high-speed mesh modification method that utilizes regularly arranged grid points and coordinate transformation. In the proposed design system, the user can immediately observe the analysis result of the design change. Furthermore, if the proposed system is executed in cooperation with the projector, designers can change the design of the motor with their fingers and discuss it on the board. Furthermore, since this system can be executed on a Web browser, it can also be executed on a smartphone or tablet PC. [ABSTRACT FROM AUTHOR]

This study presents a novel two-step optimization method that incorporates explainable neural networks into topology optimization. The deep neural network (DNN) is trained to infer the torque performance from the input image of the motor cross section. The sensitive region that has a significant influence on the average torque is extracted using gradient-weighted class activation mapping (Grad-CAM) constructed from the DNN. Then, the optimization with respect to the torque ripple is performed only in the incentive region with little influence on the average torque. The proposed method is shown to increase the average torque of an interior permanent magnet (IPM) motor by 14% and reduce the torque ripple by 79% compared with the original model. [ABSTRACT FROM AUTHOR]