Distributed Secondary Control Design Based Consensus Algorithm For Islanded Microgrids

Distributed generations(DGs)based on renewable energy have been vigorously developed in order to alleviate the contradiction between energy supply and demand and reduce the negative effect on the environment.However,the randomness and intermittentness are the main characteristics of DGs.When DGs are connected to the power grid,the power flow distribution will be changed,harmonics will be introduced.That will cause the grid to fluctuate and affect the system security and stability,etc.Microgrids allow multiple DGs to be accessed,with small investment and varied operation modes.Then,microgrids can realize regionalized power generation and power supply,mitigate the contraction between DG and power grid and improve energy efficiency.When a microgrid is in islanded mode operation,without voltage and frequency support by power grid,DGs have to rely on their own controllers to maintain the stability and power balance of the entire MG voltage and frequency,which will increase the control complexity.Therefore,it is essential to do the research on the optimization and control methods of the islanded microgrid for economic,reliable and stable operation.However,the inherent defect of droop control is that the deviations of frequency/voltage amplitude is inevitable,and power sharing performance degrades with variations in system parameters and transmission line impedances in MG systems.To overcome the drawbacks,secondary control is employed to restore the voltage and frequency to their nominal values and realize accurate power sharing.The traditional secondary regulation is implemented in the one-to-many communication mode with microgrid central controller(MGCC).It does not support plug-and-play,and the flexibility is relatively poor and the single point failure is prone to occur.Distributed coordinated control(DCC)based on sparse communication has the advantages of few information transmission,flexibility and robustness,and can solve the shortcomings of traditional centralized control.Therefore,many works foucus on the research of DCC in recent years.DCC based on Multi-Agent System(MAS)consensus theory has become one of the most important research directions.In this paper,the distributed secondary coordinated control for the islanded microgrid is the main research content,with the objectives of DG voltage,frequency control and power allocation.The main contents are summarized as follows:1)The multi-inverter parallel control structure with double closed loop for islanded microgrid is established,which is consisted of virtual impedance,current feedback and voltage decoupling loops.The impedance model and circulation current analysis are carried out,and the effect of virtual impedance and circulation current on the reactive power sharing are analyzed.Meanwhile,robust droop control has been redesigned and analyzed.Finally,the advantages of droop and robust droop control methods are verified by simulation and experiments.It is shown that secondary control is necessary to ensure the accuate control and higher power sharing accuracy.2)A power allocation scheme based on multi-agent system consensus theory(MAS)is proposed for islanded AC microgrids.The consensus algorithm is applied to achieve accurate frequency control,intermediate AC side output voltage,point of common coupling(PCC)voltage accurate control,and power sharing.Meanwhile,the AC side output voltage is operated within the error range.The secondary controller is nested on the local controller without additional energy management system(EMS)or MGCC.It only depends on the adjacent DG nodes.Information interaction realizes distributed coordinated control of islanded microgrids.Meanwhile,a small-signal model with relative few state variables.Taking 3 DGs as an example,the underlying droop control contains only 9 state variables,and the number of state variables is 25 after adding the secondary control.The small-signal model is more suitable for analyzing the influence of the secondary controller parameters on MG system stability.3)The minimum boundary of the coupling weight in the consensus theory is related to the non-zero minimum eigenvalue of the Laplacian matrix of the communication network,which is a global information.If the size of the network is very large,the calculation of eigenvalues of the network coupling matrix may be difficult to implement.And it is impossible that every DGs has the global information when there is no MGCC/EMS.In this paper,we propose a decentralized adaptive pinning control scheme for general complex dynamical networks synchronization without requiring global information of the networks,which introduces local adaptive strategies to coupling strengths and feedback gains.Then,the network can synchronize with weak coupling strengths and small feedback gains by introducing this local adaptive strategy on each node.However,the adaptive pinning control scheme is not suitable for the multi-disturbance situation,which will accumulate the coupling weight and even cause the intergral saturation.Therefore,an improved strategy is proposed to adapt to the load changing conditions for the islanded MG.4)In order to enhance the reliability of stand-alone microgrids,several neighbouring single micorgrids have been coupled to be multi-microgrids(MMGs)and exchange power in a certain region with common benefit,which is an effective way to reduce the load shedding amount and the capacity of DGs,improve power supply reliability.In this paper,a multi-microgrid with the radial network and a single bus is taken as the research topic.A multi-layer control structure and complex communication network are inevitable,which increase communication delays and complicate the control structure.To solve this problem,this paper introduces a distributed secondary control based on cluster consensus of inhibitory coupling with a power limit scheme.Compared with the inter-cluster communication coupling structure,the communication channels of the proposed scheme are minimized to simplify network topology,decrease the communication delays and control dimensions.The proposed control strategy reaches multi-valued consensus states for accurate active/reactive power sharing in different MGs.Furthermore,active/reactive power overloading in a MG cluster is discussed under an inhibitory coupling configuration.Based on the power limit scheme,each distributed generator(DG)of the overloaded MG is controlled to generate power within the rated power range for high local autonomy.In addition,an adaptive virtual impedance control is introduced to decrease voltage deviations between DGs.The analytical results are verified by a simulation in MATLAB/Simulink.