Research On Expansion Planning Of Medium Voltage Distribution Network With Distributed Generation

With the continuous increase in power consumption and the gradual increase in the number of power consumption points,it is an indispensable work to carry out expansion planning on the basis of the existing distribution network.In addition,the continuous access of distributed power sources has changed the power flow characteristics of distribution network,which has a non-negligible impact on the economic operation of the distribution network.However,due to the constraints of property rights and natural environmental conditions,the power supply company’s expansion plan for the distribution network does not include the investment and maintenance costs of distributed power generation into the planning.This article is to get the smallest construction cost from the perspective of power supply company in distributed network,considering the expansion planning of the distributed power,the smallest line loss and meeting all power supply needs.Based on the actual situation of the township distribution network,this paper proposes a distribution network expansion planning method combining dynamic Markov-long short-term memory stacking neural network load forecasting,source-load Monte Carlo simulation,and improved genetic-particle swarm algorithm.Considering the different scale of historical load data owned by each load node in the same distribution network,long and short-term memory units are used to process historical load data of different time spans.This paper divides them into three categories: daily,monthly,and annual scales.Considering that each forecast will inevitably produce errors,the dynamic Markov model of error is introduced,and then the error calculation of each influencing factor that affects the accuracy of the load forecasting is carried out to make the load forecasting more accurate.On the basis of the load forecasting,Monte Carlo method is used to simulate the load changes from the planning base year to the planning level year.At the same time,the output of the distributed power generation is simulated according to the plan of the distributed power installation report.The simulation results of load and power output are used as the basic data for the power flow calculation of the distribution network to be planned,and the improved forward push-back method is used.An expansion planning model based on the minimum construction cost of the distribution network expansion,the minimum operation and maintenance cost and the minimum line loss is proposed,and the improved genetic-particle swarm algorithm is used to solve the problem.The line loss of the distribution network is calculated using the wire current calculated by the power flow and the wire resistance.The improvement of the genetic algorithm is embodied in the coding,crossover and mutation operations,making it more suitable in the solution of the distribution network expansion planning model which contains distributed power sources.The particle swarm algorithm mainly calculates the optimal installed capacity of distributed power sources.The results of the expansion planning for actual distribution network which contains distributed power sources are obtained through research,and the effectiveness of the planning method proposed in this paper is proved through comparative analysis.