Research on Explicit Model Prediction Control of Pulse Current Source

Abstract

In order to improve the performance of pulse current source and to ensure the safe and stable of the system, Model Predictive Current Control (MPCC) method is adopted to pulse current source. However, there is a large amount of online calculation and high computational complexity, which is especially prominent in systems with a long and complex prediction time domain. To solve these problems, an explicit model predictive current control (EMPCC) method to reduce the online operation time is proposed. By using the state vector of the system as the parameter vector, the optimal control problem is equivalent to a multiparameter quadratic programming problem. The valid set method is used to solve the multiparameter quadratic optimization problem. The mapping relationship between the output voltage, output current and the state partition is obtained. The relationship between the state partition and the duty cycle is further deduced and stored in the lookup table. Based on the system state, the sequential search method is applied to find the duty cycle online, and the optimal control quantity at the current moment is obtained. This method pre-computes the online prediction and duty cycle calculation process offline, which solves the disadvantage of the model predictive current control method with large online computation. Meanwhile, this method can effectively improve the response speed of the pulse current source to the pulse load and improve the output current quality. The simulation and experimental results demonstrated the feasibility and validity of the proposed method.