Research On The Energy Optimization Methods Of Smart Distribution System With Source-Network-Load-Storage Coordination Considering Uncertainties

China’s “14-th Five Year Plan” and the national “Carbon Peaking Action Plan before 2030” clearly point out that it is an important approach to build an smart distribution system(SDS)with new energy as the main body to achieve the“Double-Carbon” goals.As a lot of flexible resources such as distributed new energy,flexible load,energy storage and flexible interconnected devices are accessed to SDS,the structure and operation scenario of SDS are undergoing revolutionary changes.And it has brought certain technical challenges for energy optimization and operation control of SDS.In particular,the intermittency of distributed generation output and the uncertainties brought by the increasing proportion of multi-type loads bring new challenges to the energy optimization of the distri bution network.In addition,it is the basic problem of source-network-load-storage coordination and optimization for the efficient operation of SDS,which is also the basis of multi-timescale energy optimization for SDS.It is necessary to study the energy optimization of source-network-load-storage coordination.Thus,this paper considers the uncertainties of source and load,and it takes the traditional active AC SDS,flexible interconnected SDS,SDS with multi-microgrids(MMG)as the research objects.And this paper is firstly based on the global coordinated optimization model of source-network-load-storage resources,and then it focuses on the research of the distributed new energy consumption optimization,network-side resource coordination optimization and load-side MMG cooperation optimization.This paper has formed a relatively complete theory and method of energy optimization for SDS with source-grid-load-storage coordination considering uncertainties.The main works are summarized as follows:(1)A mixed integer linear programming(MILP)model for global coordination and optimization with source-network-load-storage considering uncertainties is proposed.Firstly,a joint distribution model of correlation variables is constructed based on Copula theory,and an uncertain scenario modeling method considering source-load correlation is proposed by combining with Monto-Carlo sampling.Secondly,the linearized current injection models are presented for power source,flexible load(FL)and battery energy storage(BES)based on the linear power flow model.Then,a coordinated day-ahead scheduling model incorporating topology reconfiguration,BES optimization and load response is presen ted to minimize the total operational costs,and an extensible linear switching operations calculation(ELSOC)method is proposed to address the network reconfiguration.The simulation results show that the proposed extended linearized power flow model has high accuracy,and the global coordination and optimization of source-network-load-storage can greatly improve the operation economy of SDS.In addition,the multiple uncertainties significantly affect the sche duling strategy of power source,network,FLs and BES resources.(2)A joint optimization method of distributed photovoltaic(DPV)a nd BES of flexible interconnected SDS for new energy consumption is proposed.Firstly,an DPV-load joint temporal scene generation method is proposed to handle the source-load uncertainty and temporal correlation based on the characteristics of dynamic time warping.Secondly,it establishes the mathematical model of flexible resources of source,network,loads and BES,which contains the soft open points.Then,it presents a mixed integer linear programming model of DPV and BES joint optimal configuration,and the goal is to maximize the annual total power generation of DPV when considering the distribution network economy constraints.And it calls CPLEX in the GAMS software to solve the optimization problem.The simulation results show that the proposed joint optimization method can effectively improve DPV consumption when ensuring the investment and operation economy.In addit ion,it also shows that appropriate DPV activ e power reduction can effectively promote long-term DPV consumption.Moreover,the joint temporal scene generation method can accurately describe the randomness and correlation of DPV and load with less typical scenarios.(3)A multi-time-scale optimization method of flexible interconnected distribution system with self-energy storage based o n the grid-side resource coordination is proposed.Firstly,a multi-time scale active and reactive power joint optimal control architecture for coordinated utilization of multiple grid-side resources under "source network load storage" is proposed.Secondl y,in the day ahead stage,this paper proposes a long-time scale active and reactive power joint optimization model considering the on-load tap changer(OLTC),discrete and continuous reactive power compensation devices,reconfigurable switch es and SOP with energy storage.In the inner-day stage,it proposes a multi-objective rolling optimization model.And these two models are transformed into MILP model by using the linearization method.Then,the fuzzy sets of source and load uncertainties are constructe d based on KL divergence,and the transformation process of two-stage distributionally robust optimization(DRO)model for the day-ahead stage is given while the solution process based on column constraint generation(C&CG)algorithm is explained.The simulation results show that the multi-time scale coordination of grid-side resources with fast and slow characterist ics can effectively improve the economic benefits of SDS.In addition,compared with stochastic optimization,the proposed DRO method can take the economy and robustness of the operating strategy into account.(4)A distributed and coordinated low-carbon optimal operation method for MMG on the load-side considering the multi-type demand responses is proposed.Firstly,the cooperative operation and optimization framework for MMG with two-level carbon trading mechanism is constructed.Secondly,based on Nash bargaining theory,an MMG cooperative game op eration optimization model considering the two-level carbon trading and multi-type demand responses is established.The non-convex and nonlinear Nash bargaining problem is then transformed into two easy-solved sub problems to optimize the interactive elect ricity,carbon trading volume and trading price step by step.Then,an improved prediction-correction-based alternating direction multiplier method(PCB-ADMM)is proposed,and the DRO method is also introduced to realize joint solution of the distributed robust and distributed optimization.The simulation results show that the multi-type demand response can effectively reduce the total operating cost and total carbon emissions of MMG.The proposed two-level carbon trading model can better promote the electricity and carbon emission allowances sharing,and it can also improve the cooperative operation efficiency of MMG system and each microgrid.In addition,the distributed solution only requires a small amount of communication information,which can effectively protect the privacy of each microgrid.