Multi‐Objective Self‐Optimisation Control and Dynamic Behaviour Analysis of Permanent Magnet Contactors

ABSTRACT

The dynamic performance of the permanent magnet (PM) contactor's making and breaking process directly determines its service life. To overcome the limitations of existing whole‐displacement control methods, this paper proposes a multi‐objective self‐optimisation control strategy based on a backpropagation neural network proportional–integral–derivative (BPNN‐PID) algorithm. The strategy dynamically segments the displacement curve, couples the mechanical motion equation with the voltage balance equation and monitors the force and speed of the moving iron core in real time. The BPNN optimises the PID parameters online, suppressing the final speed and contact bounce. Simulation and experimental results validate the strategy. Compared without intelligent control, the proposed method reduces the contact bounce number by 79.0%, the bounce time by 53.3%, the final core speed by 73.5% and the arc length by 90%. These improvements significantly enhance the mechanical and electrical service life of the permanent magnet contactor.