Research On New Type Of Power System Flexibility Enhancement Considering The Coordination Of Source-Network-Load

Global carbon reduction campaign has further increased the proportion of renewable energy in the power grid,while the decline in proportion of traditional power generation has weakened the regulation capability of power grid.Therefore,the output variation and randomness of high proportion of renewable energy have become key difficulties for the stable operation of the power system.In addition,energy storage devices,heating ventilation and air conditioning,and distributed renewable energy station in distribution network also intensify the uncertainty of demand-side power consumption.It is urgent to enhance the power system flexible regulation capability to cope with the frequent imbalance of active power supply and demand.Making full use of source-grid-load cooperative operation technology and power market mechanism,which can improve proportion of renewable energy and stimulate potential regulation capacity,is an effective measure to enhance the power system flexibility.In this paper,the existed definition of power system flexibility is extended,which means that some operation indexes in power system normal operation and post-fault operation are included in flexibility indexes.Furthermore,the evaluation method of power system flexibility index is proposed.Then,reinforcement learning(RL),uncertainty modeling and data-driven techniques are respectively utilized for the study of RL based solar-storage station optimal capacity scheduling considering the flexibility in source side and the study of optimal energy management of demand side virtual power plant with the demand side flexibility and users’ privacy protection.Finally,combining with flexibility enhancement from source and demand side,flexibility enhancement strategies and optimal operations of power system normal state and post-fault state are proposed considering source-network-load coordination.The main research contents and achievements are as follows.Aiming at the connotation and index evaluation method of power system flexibility,this paper first summarizes the current various definitions of power system flexibility,and based on the five types of characteristics contained in flexibility,the connotation of power system flexibility has been expanded.Then,the calculation method of flexibility evaluation index for normal operation and post-fault operation of power system is proposed.Finally,an optimal operation model considering flexibility improvement measures is constructed based on the flexibility test system.The common source-network-demand-side flexibility improvement measures are used to prove the validity and feasibility of the proposed evaluation index.The participation of photovoltaic power stations with energy storage(PV-ES)in the power market can provide more regulation capacities for the power system.This paper investigates the relevant policies of the ES to allow the provision of power spot products,and an internal ES capacity scheduling strategy is proposed for PV-ES that provides three different services.Additionally,the optimal capacity scheduling model for PV-ES to participate in the power spot market is constructed.Since the model contains uncertain factors such as various market prices and frequency regulation signals,the model presents non-convex characteristics.In this paper,the reinforcement learning technology is used to solve the optimal capacity scheduling model of PV-ES,and the optimal capacity scheduling method with flexibility improvement strategy in the source side are obtained.Aiming at the problem of demand-side energy management and scheduling considering users’ privacy protection for demand-side flexibility enhancement,this paper proposes a hierarchical energy management framework that considers the equipment layer,terminal aggregation layer and demand-side virtual power plant(DS-VPP)energy management layer.In order to reduce the scheduling burden of the energy management layer,a power aggregation model of terminal aggregation layer based on the inner-box estimation method is proposed.Furthermore,the optimal energy management model of the DS-VPP is constructed,and the stochastic-ranking-method based data-driven evolutionary algorithm(SRK-DDEA)is proposed.Finally,the energy management of the DS-VPP with users’ privacy protection is achieved and the flexible ramping products from demand side are provided for the independent system operator.Lastly,based on the proposed flexibility connotation and combining with the flexible regulation capacity provided by source and demand side,flexibility enhancement strategies and optimal operations of power system normal state and post-fault state are proposed considering source-network-load coordination.Aiming at the uncertainty from the source and demand side,an uncertainty modeling method based on the combination of robust optimization and stochastic programming is proposed.Aiming at the calculation of post-fault steady-state voltage and frequency,the calculation formula of variation between pre-fault and post-fault voltage and frequency is constructed by using the static frequency characteristics of generators and electrical loads.Finally,the numerical cases verify the effectiveness of various source-load-storage regulating resources in flexibility enhancement of the power system normal operation.Meanwhile,the cases indicate the importance of ES to the flexibility enhancement of the power system post-fault operation.In addition,power system flexibility enhancement strategies and operation for normal state and power system post-fault flexibility enhancement strategies by rescheduling strategy and flexible resource planning scheme are obtained.