Distributed Coordination Control System For DC Multi-microgrids

Recently,with the unprecedented growing penetration of renewable energy sources(RES),the concept of microgrids(MGs)becomes promising anticipation to ensure efficient and resilient electric networks.Moreover,DC MGs have gained increased attention due to their inherent advantages such as better efficiency and simple controllability compared to AC MGs.Furthermore,interconnecting nearby MGs to each other forming a DC MGs cluster improves the overall reliability and availability owing to the exchanged power between MGs.The proper coordination control system is required to ensure the stable and economic operation of the MGs cluster.The hierarchal control system in DC MGs consists of primary and coordination control layers usually implemented in a centralized,decentralized,and distributed fashion.Distributed control scheme offers improved system reliability,and scalability with a reduced communication network costs compared to a centralized scheme.Also,it provides better system performance than the decentralized one.Therefore,distributed decision-making algorithms have acquired a significant interest in power management for DC multi-MGs.However,many of the existing works study equal load power-sharing among interconnected DC MGs,which might not be appropriate for MGs with various DGs having different operating costs.Consequently,the economic operation of the cluster should be considered to minimize the global total generation cost(TGC).This thesis addresses fully distributed control schemes that are functioning to optimally schedule the output power of all DGs within the DC multi-MGs.Firstly,a fully distributed hierarchical coordination controller is developed to minimize the global TGC for a cluster of DC MGs through handling the optimal power-sharing between MGs.It consists of local and global tertiary controllers,in which a two-layer communication network is modeled to share information with a reduced complexity as one or some distributed generators(DGs)of each MG are pinned from the local network to broadcast the information between MGs at the global network.Under this control structure,the global tertiary controllers(cluster-control layer),based on the linear dynamic consensus technique,generate the global optimum incremental cost(IC)reference to attain the entire cluster’s economic operation.Moreover,using the pinning-based leader-follower consensus algorithm,the local tertiary controllers assign MG average voltage reference at which the ICs of all DGs are matched at the global reference to ensure optimal operation of the multi-MGs with considering DGs generation boundaries.Finally,the nominal voltage setpoints of each DG can be determined by the secondary controller to guarantee the generation-demand power balance.The feasibility of the proposed controller is demonstrated by simulation,HIL,and experiment under various test scenarios.However,the linear consensus algorithm realizes an asymptotical convergence speed with undefined convergence time,which is unsuitable for MGs that have a fast intermittent operation of RES and frequently changing loads.Consequently,the finite-time consensus protocol is developed to attain faster convergence within a finite settling time compared with the asymptotical consensus algorithm.A novel finite-time-based control strategy is proposed to optimize the power flow among interconnected DC MGs respecting the equality and inequality constraints of the economic dispatch problem(EDP).Based on the Lyapunov analysis,the stability and convergence of the closed-loop system are analyzed thoroughly.Furthermore,battery energy sources(BES)become a dominant part of the future power system guaranteeing MG generation quality;accordingly,the economic charging/discharging of BES in the cluster is taken into consideration to increase the energy arbitrage.The proposed controller includes two control layers labeled cluster-and MG-controller.Wherein cluster controllers equalize MGs’ ICs at the global optimal value to optimize the power flow among MGs,in finite-time manner,by tuning MGs’ voltage references.MG-controller allocates DGs’ output powers economically and regulates the average voltage across the MG at the value assigned by the cluster controller.Simulation and experimental studies illustrate the effectiveness of the proposed control strategy.Regardless of the fast convergence rate for the finite-time protocol,the upper bound of the settling time is a function of the initial states,which may be unavailable,especially in large systems such as multi-MGs.Consequently,the fixed-time consensus protocol is utilized to solve the EDP for DC multi-MGs within a fixed settling time.The key features of the proposed fixed-time control are the fast convergence rate,and the maximum limit of the convergent time is irrelevant to the initial values as compared to linear consensus protocol and finite-time scheme,respectively.The convergence of the proposed controller is confirmed by rigorous analysis.The proposed cluster-control layer adjusts MGs’ voltage references for the economic scheduling of the power flow among MGs.Based on the fixedtime consensus protocol,matching MGs’ ICs reduces the global TGC through rerouting the power demand from the high-cost MGs to the lower-cost ones.MG-controller ensures that local DGs have a similar IC to economically supply the required demand and simultaneously restores MG’s average voltage at the desired reference voltage defined by the cluster-control layer.Consequently,the optimal operation of DC multi-MGs can be solved within a fixed settling time concerning line losses and the equality and inequality constraints.Finally,experimental and simulation results prove the superiority of the developed control scheme with different case studies.