Current Control of EAST Fast Control Power Supply Based on BP Neuron Network Predictive PI

One of the most important performance indexes of the EAST fast control power supply is the dynamic response speed of the output current. In view of the shortcomings of traditional BP neural network, which requires a large amount of computation and slow convergence speed, a predictive PI control algorithm of BP(Back Propagation) neural network is proposed, which can further improve the response speed of output current. The branch current prediction model of EAST fast control power supply is established. Based on linear PI control and BP neural network structure, the current prediction model is used as the adaptive objective function. The piecewise linear activation function is used to simplify the computation process and speed up the convergence. According to the error between the predicted current and the reference signal, the learning rate of BP neural network structure and the activation function factor of output layer are adjusted in real time. The steepest gradient descent method is used to realize the adaptive setting of PI control parameters. Experimental results show that the proposed control algorithm has faster dynamic response than the traditional PI control.