Research on Excitation Current Control System of the 50 kA Superconducting Transformer

Superconducting transformer, which used for power supply of the large current superconductor, is faced with the conductor sample current attenuation problem since the existence of joint resistance. A solution is proposed to solve this attenuation problem to make the conductor sample current controllable. The adaptive Proportion Integration Differentiation (PID) control strategy based on learning rates (η) Radial Basis Function (RBF) neural network (NN) is an improvement over the conventional PID algorithm. The algorithm is applied to obtain the optimal parameters for the PID controller to the working variations arising from transformer nonlinear dynamics. The controller system is carried out on the superconducting conductor test platform of the Institute of Plasma Physics, Chinese Academy of Sciences (ASIPP). The results showed that the overall deviation of the system was within 0.2%. The performance of the proposed approach is validated by superconductor experiments under practical conditions.