Research Of Hybrid Coordinate Control And Active Power Optimal Strategy For Micro-grid Cluster

A microgrid has flexible operation mode and multiple control strategy,which has the ability to enhance the reliability and power quality of electrical energy supply.Most of micro sources in a mcirogrid are interfaced through power electronic converters.The interfaces of converters increase the flexibility of control but reduce the inertia of the microgrid.The microgrid has some common technical problems like power randomness,power balance adjustment,safe and stable control and power quality when it lacks the support of large power grid.Multi-microgrid is a cluster formed by the interconnection multiple microgrids in a regional distribution system,which has specific functions and operational objectives through energy exchange.Multi-microgrids can adapt to the dynamic characteristic of large scale distributed generation and improve the robustness of system based on regional autonomy acceptance and wide-area peering.Therefore,the coordinated control strategy and optimization mechanism of multi-microgrids is an important task.This thesis concentrates on a coordinated control strategy under islanded operation and optimalized strategy under grid operation for inverter-inerfaced multi-microgrids.The main jobs of the thesis are as the following.Firstly,this thesis researches the power management strategy and coordinated control strategy of microgrid under islanded operation.Considering the characteristics of distributed generation diversification,this thesis proposed a power management and hybrid coordinated control strategy.In the hybrid control strategy,the energy storage and photovoltaic operate under master-slave coordinated strategy in order to obtain the maximum energy utilization,on the other hand,the droop unit operates according its droop characteristics.In the power management strategy,the master-slave control unit and droop unit operate alternately according multi-segment P/f characteristic curve of microgrid to achieve decentralized power management.In battery storage charging scenario,a charging amendment strategy was proposed to reduce PV output power when the SOC of battery reach its limit.Finally,the simulation results verify the effectiveness of the proposed power management and control strategy.Secondly,this thesis research the hybrid coordinated control strategy and power management of islanded multi-microgrids based on microgrid control and management strategy.In the management strategy,the multi-microgrid managed the power balance adaptively follow the multi-segment power/frequency characteristics which grouped by the power/frequency characteristics of each microgrid.In the hybrid control strategy,multi-microgrids were controlled under master-slave or peer to peer coordinate strategy according to P/f characteristics of multi-microgrids.In peer-to-peer control mode,in order to analyze the problem of frequency stability,the corresponding small-signal model is established based on the average frequency method to analyze and determine the droop coefficient.The simulation results verify the effectiveness of the proposed power management and control strategy.Thirdly,this thesis proposes a microgrid power optimal dispatch scheme by using Markov decision process(MDP).To decrease the exchange power of PCC and smooth the fluctuations in different time scales,the reward function of MDP is established as entropy.During the decision process the optimal policy and optimal entropy value are solved through policy iteration.In the policy iteration,a specific policy set of wavelet packet-probability fuzzy has been proposed.It is used as the mapping from the state set of state of charge(SOC)of energy storage to action set of charge/discharge power of hybrid energy storage.The performance of this optimal method is discussed by study cases in the end.Finally,this thesis presents a microgrid power optimal dispatch scheme by using markov decision process(MDP).To decrease the exchange power of PCC and smooth the fluctuation in different time scales,the reward function of MDP establish as entropy.The decision process gets the optimal policy and optimal entropy value through policy iteration.In the policy iteration,a specific policy set of wavelet packet-probability fuzzy has been proposed,which used as the mapping from state set of state of charge(SOC)of energy storage to action set of charge/discharge power of hybrid energy storage.The performances of this optimal method are discussed by some study cases in the end of this paper.