Distributed Control Methods For Microgrids Considering Communication Constraints

In order to alleviate energy crisis and improve environmental condition,the efficient utilization of renewable energy becomes important and it has been a developmental trend.A microgrid,which can effectively manage the distributed generators(DGs)and enhance the penetration of renewable energy,provides users with flexible and diverse energy.Microgrids,which lacks of voltage and frequency support from the main grid,exhausts to deal with fluctuation of renewable DGs’ outputs and stochastic load changes.Thus,it draws more focus on its control strategies.Furthermore,the communication delay and the packet losses in communication system will have a negative impact on the working performance of the microgrid.Therefore,it is necessary to analyze and study the control strategy of the microgrid when considering constrained communication.Based on multi-agent system and considering the impact to the microgrid of the communication delay and the packet losses,this paper proposeda distributed control method of microgrid when communication is constrained.First,a two-layer distributed control model for microgrids was established and a fast distributed control law was designed.Microgrids can work stable and the system power could be dispatched among controllable DGs according to their capacities if the outputs of controllable DGs are regulated by Agents in term of control laws when considering bounded communication delay and packet losses.Secondly,a distributed control law contained power estimation term was extended,which could effectively promote control performance when there were large-scale packet losses.Finally,the MATLAB/Simulink digital simulation system is built to test the proposed distributed control strategies.The main jobs of the thesis are as following.(1)In this paper,a two-layer distributed control model for microgrid was proposed,in which the top layer was the communication network consisted of Agents and the bottom layer was the microgrid composed of distributed generators(DGs)and loads.First,the topology of communication network was introduced.Then,from any given communication network,a set of distributed control laws werederived for Agent,which can ensure the system power dispatched among controllable DGs according to their capacities.Moreover,it is proved that the system power demand-supply balance is guarantted and proportional outputs of DGs are achieved iteratively,if agents apply the derived control laws.Finally,the proposed distributed control strategy is tested on the simulation system.The simulation results show that the microgrid can work stableand active and reactive power can be dispatched proportionally even when there were the communication delay less than 30 ms,the probability of packet loss under 0.4 and the communication link-broken.(2)In order to eliminate the negative impact to the microgrid brought by large-scale packet losses in communication system,this paper utilized the historical information to estimate the power change and proposed a distributed control law which contained power estimation term.Firstly,a communication protocol for Agent was designed,which could reduce the communication numbers among Agents.Secondly,Agents used the saved historical information to estimate power change,when there were packet losses.Thus,for any given communication network,a distributed control law which contained power estimation term was proposed.Furthermore,the convergence of the control laws with power estimation term is anlaized theoretically.Finally,four simulation cases were designed to evaluate the performance of microgrid when considering large-scale packet losses.And the results show that the system voltages and frequency vary in a normal range and the proportional outputs of controllable DGs is achieved when the probability of packet losses is high.(3)The MATLAB/Simulink digital simulation platform of microgrids was built in order to test the distributed control strategies.First of all,four different types of DG simulation model including micro-turbines(WTs),PMSG-WTs and PVs and battery energy storage system(BESS)were established.Later,different control models of DGs were debugged and the microgrid with different communication networks was constructed.Moreover,the simulation models of communication delay,packet losses and link-broken were established.Finally,the control performance of the distributed control strategy considering communication constraints is verified by the digital simulation paltform.