Power Coordination Control Of Microgrid Based On Multi-agent Consensus

The microgrid has become an important part of "new power system",which can realize the renewable distributed generation(DG)of high penetration to connect to the grid and transmit the electric energy.However,with the increasing number of DGs such as photovoltaic power generation and the access of a large number of high-power loads,the operation stability and economic profitability of the microgrid will be affected.In order to achieve the goals of dynamic balance,economic optimization,reasonable distribution of the power,frequency and voltage optimization on the "source" side and "load" side of the microgrid,it is necessary to design the strategy of power coordination control and research the distributed hierarchical control architecture and distributed control algorithm from the perspective of the "source" side and "load" side of the grid-connected microgrid,so as to optimize the active output between the DGs of the "source" side as well as the grid and economically allocate the charging power of electric vehicle(EV)loads participating in demand response on the "load" side.In the meantime,the goals of power coordination,economic distribution and the balance of supply and demand with respect to the output power among DGs of the "source" side and the response power among EV clusters of the "load" side are realized.Accordingly,it is imperative to study and propose the power coordination control technology of the "source" side and "load" side of the microgrid with the practical engineering application value.Therefore,this thesis makes the following works:1)First of all,the control strategy and primary control methods of the microgrid as well as the control approaches of DG inverters are introduced.The principles and characteristics of each control strategy and control method are analyzed,with the exposition of the importance of studying the power coordination control method of the "source" side and "load" side of the grid-connected microgrid.At the same time,the consensus problem and principle of multi-agent system(MAS)as well as the features of different consensus protocols are researched.2)Secondly,the architecture and working mechanism of gridconnected microgrid system as well as the principles of DG participating in power coordination control on the "source" side and EVs participating in response power coordination control on the "load" side are expounded.By employing the attributes and characteristics of agents,on the basis of the control objects on the "source" side system and "load" side system,the MAS is constructed.Simultaneously,the economic optimization cost model of DGs of the "source" side and the demand response cost model of EVs of the "load" side are established.3)In view of the problems of power incoordination distribution and higher operation cost on the "source" side of the grid-connected AC microgrid with multiple DGs,on the foundation of the economic optimization cost model,an improved pinning consensus algorithm based on incremental cost rate(ICR)is proposed.The fast convergence of the improved algorithm is utilized to achieve the rapidity of economic optimization response of the "source" side and the minimization of total operation cost of DGs.Subsequently,a method of DG participating in power coordination control on the "source" side of grid-connected microgrid based on improved pinning consensus is proposed.The coordination of the economic optimization operation control layer,the optimization control layer of frequency and voltage as well as the droop control layer is employed.As a result,the economic distribution and rapid responsiveness of active output of each DG on the "source" side,the balance of supply and demand,the fast recovery of frequency and voltage of point of common coupling(PCC)as well as the coordination stability operation of DGs and the grid are realized.In the end,the theoretical analysis and simulation research are carried out on the main factors affecting the power coordination at the "source" side.The conclusion is drawn that the operation state of the "source" side system cannot be optimized due to the change of the grid electricity price and the plug-andplay of DG,but the "source" side system can still reduce the total operation cost and satisfy the total demand of load power consumption.The total operation cost of the "source" side will be augmented by increasing the parameters of the operation cost model.Also,the stability,dynamic control performance,economic cost optimization degree and power coordination control precision of the "source" side system will be reduced due to the unreasonable settings of protocol parameters and communication topologies as well as greater time-delay.4)Aiming at the problem of multiple and higher power load disturbance on the "load" side,a typical EV load that participates in demand response is served as an example in this thesis.On the basis of the demand response cost model,an improved leader-following consensus algorithm based on ICR is presented,which quickly eliminates the dynamic deviation of total power on the "load" side.Thus,the balance of supply and demand of power based on demand response as well as the minimization of total response cost on the "load" side are completed.Afterwards,a method of EVs participating in demand response power coordination control on the "load" side of grid-connected microgrid based on improved leader-following consensus was designed and the distributed hierarchical control architecture on the "load" side was constructed.The EV charging stations in the EV response layer are employed to carry out the demand response commands to achieve the coordination operation among EV clusters,economic distribution of response power,peak shaving and valley filling as well as the recovery of PCC voltage of charging stations.Moreover,the proposed control method of the "load" side is characteristic of rapid response capability and better steady-state performance.Finally,the main factors influencing the power coordination on the "load" side are analyzed and simulated.The conclusion is drawn that the response state of the "load" side system cannot be optimized on account of the plug-andplay of load agents and the alteration of the grid electricity price,but the total response cost can still be decreased and the total task of load response also is completed by the "load" side system.In addition,the total response cost of the "load" side will be risen by the increase of load model parameters.The response power regulation performance and response cost optimization degree of the "load" side system will be reduced due to the unreasonable settings of protocol parameters and communication topologies.The precise control of response power will not be attained.