Design Of Intelligent DC-DC Converter Based On FPGA Control

Digital DC-DC converters have become an important part of the power supply field due to their high efficiency and low temperature rise,and they are gradually replacing low-power linear power supplies.With the development of power technology,DC-DC converters put forward higher requirements on the performance of their control algorithms,which led to the gradual replacement of traditional PID control algorithms.As a basic unit of neural network,single neuron has simple structure and possesses the learning ability of neural network.The intelligent PID algorithm which is formed by combining single neuron and PID can effectively compensate for the deficiency of traditional PID control.Due to the parallel operation of FPGA,it is suitable for designing intelligent algorithms with lots of complex calculations.In recent years,the status of FPGA in DC-DC converter controllers has gradually consolidated.This thesis studies and designs an improved single neuron adaptive PID control Buck DC-DC converter based on FPGA.The main work is as follows:1)The working principle of the Buck circuit in DC-DC converter is analyzed.The Buck power stage is modeled by state space averaging method.The transfer function is derived and the Buck power stage is simulated by MATLAB.Then according to the simulation and design requirements the appropriate components are selected to build a hardware experimental platform.2)By studying the structure and principle of traditional PID,the traditional PID control's shortcoming is pointed out.Then the single neuron adaptive PID control algorithm is proposed.At the same time,the expert control and single neuron algorithm are combined to form an improved single neuron adaptive PID control algorithm.Then the simulation files of the three algorithms are written in MATLAB and the comparative simulation research is carried out.The simulation results show that the improved single neuron adaptive PID control algorithm has faster response speed and better stability.3)Using FPGA as main controller for DC-DC converter and using Quartus II as the development platform for software development of FPGA.According to the set index and simulation results,the appropriate FPGA data type and operation mode are selected.The whole Buck circuit software system is divided into AD sampling module,DPWM module and intelligent PID algorithm module,and the above three modules are realized by using Verilog HDL.4)Based on this,the steady state and dynamic performance tests of the designed Buck-type DC-DC converter are carried out.At a switching frequency of 200 kHz,a typical input voltage of 12 V,an output voltage of 3.3V,and a rated current of 1A,the test results of the system are as follow: The output voltage is stable at 3.342 V,the steady error is about 1.3%,the ripple factor is 1.2%,the system power on setup time is about 250?s,the startup process is basically no overshoot,and the recovery time is 150?s in case of sudden load change,the dynamic and steady state performance of the power supply have reached the set target.