Multi-timescale Reactive Power Coordination Optimization Method For Distribution Networks Considering The Regulation Capability Of Distributed Generation

With the improvement of distributed power generation technology and the reduction of its power generation cost,the penetration of Distributed Generation(DG)has been further increased in the distribution network.Due to the uncertainty of the power emitted by DG,when a large number of them are connected to the distribution network,they may have a negative impact on the voltage quality of the distribution network,and increase the difficulty of the optimal reactive voltage of the distribution network.Therefore,on the basis of understanding the influence of DG on the operation characteristics of the distribution network,considering the capability of DG to participate in reactive power and voltage optimization,and formulate a reasonable reactive power optimization strategy,which has important theoretical and practical significance for the safety and stability of distribution networks with DG.Firstly,the influence degree of DG on the operating characteristics of the distribution network is analyzed.From the perspective of qualitative and quantitative,,using theoretical derivation and establishing index system,the impact of distributed power sources on the distribution network voltage,network loss,and power flow is analyzed.Then,different connection scenarios of distributed power are set up,the influence law of DG on the operation characteristics of the distribution network is obtained by simulation when connecting the distributed generation in different locations and capacities.Secondly,choosing photovoltaics as a typical representative of DG,considering that the distribution network can measure and control photovoltaics,a multi-time scale reactive voltage optimization method for the distribution network involving the photovoltaic multi-state regulation capability is proposed.By studying the ability of photovoltaic to participate in reactive power optimization of distribution network under different states of output powers,a multi-state regulation model of photovoltaic is constructed.Based on this model,combined with other reactive power regulation equipment in the distribution network,a multi-time scale reactive voltage optimization model is established.In the day-ahead stage,the action plan of slow action equipment is formulated based on the optimization of prediction information.In the intra-day stage,based on the results of day-ahead optimization,the output state of each photovoltaic is coordinated and optimized to minimize the impact of photovoltaic and load power prediction error on power quality of distribution network.After that,the effectiveness of this method is verified by simulation with specific examples.Through the multi-time scale reactive voltage optimization method proposed in this thesis,the full utilization of photovoltaic output is realized,and the power quality of the system is also significantly improved.Finally,taking the distribution network with large-scale access of small-capacity photovoltaic as the research object,and based on the actual situation that the grid can measure but cannot control photovoltaic,a reactive voltage optimization method for photovoltaic participation is proposed under the condition of incomplete information.First,the photovoltaic cluster control model is established through partial information of photovoltaic side measurement.Then the virtual distribution network equivalent network is established by means of sharing information between photovoltaic clusters.Then,with the voltage security and stability of the virtual distribution network as the goal,the established reactive power optimization model of the virtual distribution network is solved through the cluster control system of the photovoltaic side.According to the obtained optimization scheme,the output of each photovoltaic cluster is adjusted,and then the power and voltage distribution of the distribution network is affected.Finally,a specific simulation example is used to verify the feasibility of the proposed method.Starting from the photovoltaic side,this method uses the virtual distribution network to better solve the reactive power and voltage optimization problem of the distribution network,and overcomes the difficulty that the distribution network cannot achieve long-distance centralized regulation when the contact information between the source and network is incomplete.