Structural Optimization Design of Vacuum Interrupter Contact Based on Multi-objective Grey Wolf Optimization Algorithm and RBF Neural Network
A suitable magnetic field distribution of the vacuum interrupter contact when breaking is conducive to im-proving its interrupting performance.At the moment of closing,the electrodynamic repulsion force between the moving and static contact can cause the contact to bounce.To solve the above problems,this paper firstly establishes a three-dimensional model of cup-shaped longitudinal magnetic contact with iron core,and carries out the calculation of magnetic field distribution and electrodynamic repulsion force.In order to further improve the contact performance,the RBF neural network model is constructed with the contact plate slot length,slot width,radial deflection angle,the heigth of cup holder slope slot,and individual slope slot up-and-down rotation angle as inputs.The maximum longitudinal mag-netic field strength in the center plane of the contact gap at the peak current moment,the hysteresis time at the center point at the over-zero moment,and the electrodynamic repulsion force between the dynamic and static contact at closing are taken as outputs,respectively.Finally,the structure of contact is optimized by combining the RBF neural network model with the multi-objective grey wolf optimization algorithm(MOGWO).The results show that,compared to the ini-tial structure,when the length of the contact plate slot is 19.74 mm,the width is 3.94 mm,the radial deflection angle is 19.9°,the height of the cup holder slope slot is 18.0 mm,and up-and-down rotating angle rotation angle is 119.2°,the contacts have a better magnetic field characteristics and the electrodynamic repulsion force between the dynamic and static contact is significantly reduced,which is conducive to improving the working performance of the contacts.
vacuum interrupter contactelectrodynamic repulsion forceRBF neural networkmagnetic field character-isticsmulti-objective grey wolf optimization algorithm
万方数据
维普
