Design Of DC-DC Converter And Its Active Disturbance Rejection Control

The rise of intelligent warehousing drives the development of rail-guided-vehicle(RGV),and DC-DC converter is an indispensable component in RGV.In order to make RGV have higher working efficiency,shorter charging time,stronger endurance,wider range of adaptation,put forward new requirements for its internal DC-DC converter.And this let DC-DC converter develop towards miniaturization,high power,high efficiency,high power density,strong robustness.In this paper,a DC-DC converter with4 k W output power is designed,which is studied from the aspects of circuit topology selection,basic working principle,circuit hardware design,main loop parameter calculation,controller design,etc.Finally,simulation and experiment are carried out,and a prototype is made.The main works are as follows.This article first provides a detailed introduction to several common isolation circuit topologies of DC-DC converters.The phase-shifting full-bridge(PSFB)topology as the hardware topology of the designed DC-DC converters.Then,the basic working principle of PSFB topology is illustrated,the way to achieve soft switching is analyzed,and the mathematical model of the topology is given.According to the design requirements,the main components of the PSFB topology were calculated,designed and selected.Then design the circuit of the DC-DC converter,including the main circuit,driving circuit,detection circuit,hardware protection circuit and so on,complete the hardware design of the DC-DC converter.In terms of control system design,this paper adopts the double closed-loop control strategy of voltage outer loop linear active disturbance rejection control(LADRC)and current inner loop PI control for DC-DC converter.LADRC estimates the unmodeled part of the system and external disturbances to improve the robustness of the DC-DC converter and enhance the control effect of the controller on the converter under different working modes.In addition,in order to solve the balance problem between LADRC fast estimation of total disturbance and suppression of noise amplification,reinforcement learning agent is used to dynamically adjust the observer bandwidth of LADRC.When the converter is disturbed by external disturbance,LADRC will have a large observer bandwidth to quickly estimate the disturbance,and make the system return to steady state.When the converter is in steady state output,the agent will provide small observer bandwidth to LADRC to reduce the influence of noise amplification on the converter output ripple.Finally,the state-action-reward-state-action(SARSA)algorithm is used to train the reinforcement learning agent,and the PSFB DC-DC converter model is built,and the simulation verification of the application of the designed double closed-loop controller to the DC-DC converter is completed in MATLAB/Simulink.Completed the prototype manufacturing of DC-DC converter,and the experimental results verify that the designed DC-DC converter and double closed-loop controller achieve the expected goal.