Design Of Excitation Power Control System For Superconducting Transformer

The superconducting conductor is the main component of the superconducting magnet system in the fusion reactor.The research on the performance of superconducting conductor is of great significance to the whole magnet system.The superconducting transformer system provides the required current for conductor performance testing and it is necessary to accurately control the current when the current sharing temperature is tested.Therefore,this study is based on the design of a superconducting transformer control system to conduct related research.First,it is indicated that there is a problem of continuous attenuation of the secondary current in the“flat-top” state during the conductor test of the superconducting transformer.Through the analysis of the internal working principle,a closed-loop current negative feedback control scheme is designed.The hardware design of the control system based on a compact Rio controller includes an acquisition device,control device,and control system.The algorithm of the controller adopts the intelligent algorithm of PID control based on a neural network,which overcomes the difficulty of traditional PID parameter adaptive tuning.This study puts forward improvement measures for the step size adjustment in the process of tuning PID parameters by radial based on function neural network.The step size is adjusted adaptively according to the loss function to further improve the learning efficiency of the algorithm.Simulation results show that the improved algorithm can reduce the overshoot of the system,speed up the adjustment time of the system,and has better performance.The software of the controller is written in Lab VIEW language.Through the analysis of the three-tier architecture of compact Rio controller,the programs of FPGA,real-time controller,and upper computer are written,and the data communication mode between each layer is defined.The experimental results show that the steady-state error of the control system is less than0.2%,which meets the requirements of the system design,and lays a solid foundation for the subsequent application on the 100-k A conductor test platform.