Research On Intelligent Controller Of Soft Switch DC-DC Convertor Based On DSP

When the switch works at the state of hard switching, serious EMI will occur; at the same time the switch will also endure the big stress from voltage and current, which may damage the switch. The method to solve the problem mentioned above is reducing the on-off's losses. Besides choosing the switch with high performance, another way to reduce the losses is to soft-switching the converter's switch.This paper takes the ZVT-Boost converter as the research object and adopts PWM control to achieve the constant frequency control. The assistant circuit only works when the switch functions which can reduce the losses of assistant circuit. The assistant circuit is not connected in series with main circuit. It is paralleled with main circuit so that the losses of main circuit can be reduced too. The assistant circuit will not increase the stress from main switch's voltage and current when working.Open-loop control is adopted in most soft switching converters at the present time. Although the circuit's main and assistant switch can be soft switching in the method, sometimes the stability and anti-jamming of output voltage can be easily ignored. Considering these problems, closed-loop control will be applied in the ZVT-Boost converters to ensure the performance when soft switching functions.Double closed-loop control is built on the basis of main circuit with conventional PID controller and its working principle is analyzed in this paper. The simulation model is set up and the result proves the theory. In order to get better controlling performance, the paper also analyzes the relevant theory of intelligent control-neural network combined with PID and the simulation of converter is given. The simulation result shows that the system with the neural network controller has good output precision and capacity of adapting the circumstance. According to the experiment on the board, the advantage of the intelligent control is proved once again, which expands the later research on intelligent control of soft switching.