Abstract:

In application fields such as distributed energy resource grid integration and electric vehicle charging stations, there are stringent requirements for the dynamic performance of output current. To meet the requirement of a specific product to achieve both good dynamic performance and high steady-state accuracy when outputting a slow-varying sinusoidal current, an Improved Single Neuron Adaptive PID(ISNA-PID) control strategy is proposed. This strategy integrates the Single Neuron Adaptive PID(SNA-PID) with expert control to enable online automatic adjustment of the SNA-PID connection weights. Simulation and experimental verification of both the SNA-PID and ISNA-PID control strategies were conducted. The results demonstrate that in a power supply system based on a three-phase interleaved parallel Buck converter, the use of the ISNA-PID control strategy yields significant improvements:the response speed is increased by approximately 27% and 13% compared to conventional PID control and SNA-PID control, respectively, while the steady-state accuracy is improved by approximately 76% and 60%, respectively. These results validate the effectiveness and superiority of the proposed ISNA-PID control strategy.

Key words: three-phase interleaved parallel Buck converter, single neuron, adaptive PID control, expert control