The virtual prototype model of the three-phase magnetic retaining relay was established.Based on the orthogonal test results,the BP neural network and NSGA2 genetic algorithm were used to optimize the design of the relay.ANSYS and Adams/View were used to establish the static and dynamic models of the three-phase magnetic holding relay,and experimental verification was carried out.Taking the size of the part structure of the electromagnetic system as a variable,taking the breaking current value and the magnetic holding moment after pulling as the target,five factors and five levels orthogonal experiments were carried out to establish the data set.BP neural network was used to train the dataset,and the network functions of two optimization objectives were obtained.The NSGA2 genetic algorithm is used to optimize the objective function and update the data set,and so on until the optimal solution meets the requirements.The simulation results show that the breaking time of the optimized prototype is reduced by 23.89%,and the suction time and the bounce time are reduced by 29.2% and 47.05%,respectively,which provides a reference for the manufacture of the prototype.
关键词
磁保持继电器/虚拟样机技术/正交试验/神经网络/遗传算法/优化设计
Key words
Magnetic holding relay/Virtual prototype technology/Orthogonal test/Neural network/Genetic algo-rithm/Optimization design
维普
万方数据