Research On The Optimal Capacity-allocation Of Distributed Generation And Battery Energy Storage System Of Active Distribution Network

With the development of new energy power generation technology,the scale of distributed generation system is expanding rapidly.While alleviating the traditional energy consumption at the same time,the new energy utilization rate is low,the cost is wasted and a series of problems.How to improve the utilization of distributed energy resources and ensure the reliability of power supply system to optimize the allocation of energy storage system is an urgent problem to solve in the active distribution networks.In this paper,the problem of capacity allocation of distributed generation and energy storage system in active distribution network is studied.The main research aspects are as follows:Firstly,the distributed generation,energy storage system and controlled energy efficiency load are studied.The mathematical models of distributed generation including distributed wind power generation system and photovoltaic power generation system are established.The mathematical models of energy consumption response of the controlled energy efficiency load including air conditioning,water heater and lighting load are established.The battery energy storage system is a linear mathematical model and the mathematical model of inverter is established.In view of the demand side energy efficiency load management,formulating the optimal scheduling strategy for energy efficiency load of single user combined with the time-of-use electricity price and customer satisfaction.In order to make rational use of the energy storage system,the charge and discharge management strategy of energy storage system is formulated.Secondly,in order to select the typical sample data of natural resources,using the KMO and Bartlett test to select the optimal relevance sample data for principal component analysis.The weight,scale factor and translation factor in the prediction model of wavelet-BP neural network are corrected by the momentum-adaptive learning rate correction.For the problem of optimal capacity-allocation of distributed generation and battery energy storage system in this paper is a multi-objective problem,the harmony search algorithm is used to solve this problem.Aiming at the convergence performance of the basic harmony search algorithm and easy to fall into local optimal problem,the algorithm is improved in this paper.With the search process,the memory retention probability,harmonic tuning range and fine tuning perturbation probability are used dynamic parameter update to adjust.In order to improve the convergence performance of basic harmony search algorithm,it is combined with genetic algorithm effectively.Thirdly,short term forecast of wind speed resources in summer typical day and winter typical day compared with traditional BP neural network and wavelet-BP neural network prediction model,the prediction results show that the correctness and accuracy of the improved wavelet-BP neural network prediction model based on principal component analysis.With the wind speed,solar irradiance and load demand data of the summer typical day and winter typical day in an active distribution network based on distributed generation forecasting,the multi-objective optimal allocation model includes the allocation of capacity and cost of the distributed generation and energy storage system,new energy abandoned power rate and loss of power supply probability.The improved harmony search algorithm is used to solve the problem.Finally,the correctness and effectiveness of the improved harmony search algorithm and optimal allocation model are verified by the configuration results.Finally,in order to further reduce the allocation capacity and cost,the demand side energy efficiency load optimization management is considered.The results compared with the without considering the load optimal management show that the configuration cost decreased by 7.1%,loss of power supply probability decreased by 2.8%and new energy abandoned electricity rate dropped by 2.1%.The feasibility and correctness of the single user energy efficiency load management strategy are verified by the results.The results show that it can reduce the cost of electricity consumption,further reduce the allocation capacity and cost of distributed generation and energy storage system and improve the reliability of the power supply with the demand side management to optimize the controllable load in the active distribution networks.