Application Of Intelligent Particle Swarm Optimization In Distributed Energy Planning

With the social and economic development,Chinese electricity consumption has increased exponentially,and fossil energy is limited and unevenly distributed.In addition,the large-scale use of fossil energy will bring great negative impact on the environment,so the development of clean distributed energy is imperative.However,the high penetration rate poses a great challenge to the traditional distribution network.Therefore,how to properly allocate distributed energy is of great significance for building a reliable,economical and environmental friendly distribution network.Particle swarm optimization?PSO?has been widely used in distributed energy planning due to its advantages of easy implementation,high precision and fast convergence.Therefore,this paper selects the improved particle swarm optimization algorithm to study and analyze the sizing of stand-alone micro grid systems and the sizing and siting of distributed generation.The main work of this paper is summarized as follows:Firstly,according to the working principle of distributed power generation?DG?,this paper divides it into PQ,PV,PI and PQ?V?nodes,and gives the processing methods of different node types DGs in power flow calculation.Secondly,the traditional forward-backward power flow coding method is improved,and the node layered forward-backward power flow calculation is proposed,and its superiority is verified by simulation.Then,in order to analyze the influence of DGs on distribution network,DGs are selected to join the distribution network for power flow analysis.The results show that reasonable DGs access can improve the voltage and loss of distribution network.Therefore,the necessity of DGs access in distribution network is proved.Then,an adaptive hybrid two-group particle swarm optimization algorithm is proposed,which combines the advantages of l_bestest version of search ability and g_bestest version of convergence speed.The effectiveness of this algorithm is proved by the multi-peak function Rastrigin test.Then,the case of the photovoltaic/wind/diesel/battery micro grid is carried out by using the improved load following strategy.Typical wind/light/load change curves are selected to optimize the micro grid with the objective of the investment cost of the micro grid and the highest new energy utilization rate.Finally,the optimization results of photovoltaic/wind/diesel/battery,photovoltaic/wind/diesel and photovoltaic/wind/battery are compared,under different load loss rates.The optimization results reflect the importance of batteries in microgrid and the importance of rational allocation of microgrid capacity.Finally,this paper proposes an improved multi-objective particle swarm optimization algorithm,which adopts the idea of niches sharing when selecting the global optimum,and then selects its global guiding particles according to the proportional operator.In order to prevent the algorithm converges to the Pareto frontier of local solution,this paper put forward the time-varying mutation operator,and it improve multi-objective particle swarm precocious phenomenon.And the performance of the algorithm is tested with ZDT1,ZDT3,ZDT6 and DTLZ1 standard test functions.Then,considering considering time sequence characteristics of loads and DGs,this paper established the t time sequence characteristics curves of spring,summer,autumn and winter.Finally,an optimization model is established based on the annual investment cost of DGs,the purchase cost of the main grid and the voltage deviation.Three experiments are implemented in the PG&E 69-bus system,the WT-PV combination,only WT and only PV respectively.