Research On Coordinative Grid-power Source Planning For Distribution System Considering Demand Response And Dynamic Cluster Partition

With the global depletion of fossil fuels and the continuous increasing awareness of environmental protection,the technology of generating electricity from renewable energy resources has been more and more widely used.Compared with the traditional power generation technology,the distributed generation(DG)has the advantages of low investments,environment friendly and flexible control,etc.The DG can provide an effective way for users to consume clean energy through coordination and complementation with large power grids.With the large-scale access of DG in the distribution system,its characteristics of intermittent output,high permeability and decentralization are amplified,which changes the distribution of power flow and the characteristics of the power grid,and also brings new requirements and challenges to distribution system planning.In order to reduce the impacts of the increasing DG permeability and the mismatch between DG and load on the safety and economical operation of distribution system,this thesis firstly rationally arranges the power demand of users through demand response(DR)method from the perspective of loads,which effectively reduce the load peak-valley difference and improve the matching degree between resource and loads.Secondly,the elements of "distributed generation cluster" are introduced into the distribution system planning to promote the local absorption capacity of renewable energy by improving the power complementarity between resources and loads within clusters.The main research contents of this thesis are as follows:Firstly,this thesis systematically summarizes the mathematical model and solution method of distribution system planning.Then,the concept of distributed generation cluster is introduced,and the cluster partition is summarized from the aspects of cluster partition indices,principles and methods.Then,the comprehensive performance indicators that are suitable for planning are proposed taking into account both structure and function.And the necessities of considering clusters for the coordinated planning of networks and resources are illustrated.Secondly,considering the influence of adjustable resources on the load side on the active power balance of the distribution system,this thesis uses the dynamic time-of-use pricing strategy to adjust the load prediction curve and the index of cluster partition,and further constructs the evaluation index of time-of-use price response,and makes a detailed measurement of the response effect.Based on the fixed cluster structure,thecoordinated planning model of networks and resources is established embedded with interruptable loads.The model aims at minimizing the annual costs of society.The planning contents include distribution network,distributed power supply and interruptable load.The constraint conditions include electrical constraints and network constraints.The solutions are obtained by using the commercial optimization software ILOG/CPLEX.Taking an actual distribution network as an example,the validity of the proposed model is verified by comparing the planning results under different scenarios.Finally,aiming at the problem that the cluster partition results can not adapt to the dynamic changes of DG and networks in the current planning considering cluster partition planning,this thesis establishes a two-layer optimization model that cluster partition and grid-power resource planning can dynamically match each other,which focuses on solving the problem of interactive adaptation between dynamic changes of networks and resources and cluster partition during the planning process.The network structure as well as the locations and capacities of DG are optimized at the upper-level of the planning mode with the objective of the lowest annual comprehensive costs under the given cluster partition results.Based on the new network topology and new DG configuration in the lower-layer partition model,the cluster partition is conducted again according to the division indices.The new cluster partition scheme obtained in the lower-layer is fed back to the upper-level planning.The entire optimization process is iterated,and the planning model is solved by the improved genetic algorithm.The results and discussion of related models,algorithms,and calculation examples are given in this thesis.