Research On Distributed Secondary Control Of DC Microgrid Based On Consensus Algorithm

In the hierarchical control of DC microgrid,droop control is usually adopted to achieve current or power sharing among multi-sources modules of the microgrid,with the aim of improving the stability of multiple parallel modules.However,the basic droop control adopts a constant droop coefficient(i.e.,virtual resistance),which will not only improve the accuracy of current sharing,but also lead to a large DC bus voltage drop.Thus,it is difficult to meet the balance between current sharing and DC bus voltage regulation of multiple parallel modules at the same time.Therefore,on the basis of droop control,this thesis adopts the distributed secondary control strategy based on the multi-agent dynamic consensus algorithm to raise the bus voltage and improve the accuracy of current sharing.Furthermore,in order to raise the bus voltage quickly and effectively and improve the robustness,this thesis focuses on the research on the convergence performance of the consensus algorithm under various complex communication conditions and proposes two improved algorithms,which improves the stability of the operation of the microgrid effectively.Firstly,this thesis analyzes the model of multiple parallel battery energy storage units and basic multi-agents consensus algorithm.Then,based on the traditional droop control and the adaptive droop control of the state of charge(So C)of battery,the distributed secondary control is introduced to realize DC bus voltage recovery and So C balance respectively.Whereafter,the photovoltaic power generation unit is added to simulate the island DC microgrid power generation system.In addition,on the basis of distributed secondary control,the application of the consensus algorithm in achieving the power flow among clusters of the DC microgrid is analyzed,and the feasibility of the control strategy mentioned above is verified by simulation.Secondly,in order to analyze the performance of time-consensus algorithm in complex communication cases,three typical communication faults are selected as research objects,including communication delay,input disturbance and communication link disconnection.Then the dynamic performance and steady performance of the consensus algorithm under three communication conditions are analyzed,and the simulation results under different communication conditions are compared and analyzed.Moreover,two improved algorithms are proposed to solve the problems of communication delay and communication link disconnection respectively which have great influence on the performance of the consensus algorithm.The simulation results show that the improved consensus algorithm has strong robustness.In addition,the design criterion for the weight coefficient of the discrete consensus algorithm is given,and the accelerated convergence of the consensus algorithm is realized,which improves the dynamic performance of the secondary control both in simulation and experiment.Finally,the DSP28335 digital controller and Buck/Boost interface converter are used to build an experimental platform with multiple units in parallel.Then,according to the requirements of distributed communication,the data communication program is written in code composer studio(CCS)software,and the wireless data communication is realized through ZigBee communication networks.Experimental results prove the effectiveness of the secondary control strategy based on the consensus algorithm for DC bus voltage recovery,and show the effectiveness and practical values of ZigBee wireless communication for distributed secondary control.