Research On Optmial Power Supply Scheme For Power Business Expanding With Distributed Generation

Power business expanding is the front position of power grid company.Its main task is to formulate a reasonable power supply scheme to connect customers to the power grid at the appropriate location according to customers' power demand and combined with the operation status and planning of power system.The existing problems of the current power business expanding scheme include:(1)the formulation of the scheme is not closely related to user load forecasting,especially the uncertainty of distributed power supply is not taken into account;(2)inconsistency with power grid planning.Therefore,it is necessary to formulate a reliable and economical power business expanding scheme based on load forecasting of various types of power users,especially considering the uncertainty of distributed generation.The main work of this paper is as follows.1)A medium and long term load forecasting method for various power users based on deep belief network is proposed.Firstly,the grey relational degree algorithm is used to analyze the correlation between various influencing factors and various power users'loads.Then,the fuzzy C-means clustering algorithm is used to classify the loads of various power users.Finally,based on the deep belief network,the medium and long term load forecasting models of each classification are established,and the actual data are collected to train and test the forecasting models.The test results show that the accuracy and stability of deep belief network in prediction are higher than that of BP neural network,which has practical application value.2)The influence of uncertainty of distributed generators' output on power grid and power supply scheme is analyzed.Based on the historical output data of different types of distributed generators,the probability distribution model of distributed generators' output is established,and the distribution of node voltage and power flow in power system is calculated by probability power flow calculation based on Latin hypercube sampling.The test results of numerical examples directly and clearly reflect the operation status of power grid before and after the distribution generators are connected.3)Considering the uncertainties of the output of distributed generators,considering the load forecasting of various power users,aiming at minimizing the total economic cost in the planning stage,and satisfying the constraints of ensuring the safe and stable operation of the power grid,an uncertain planning model of power supply scheme is established.The probability distribution of node voltage and power flow in power grid is calculated by probability power flow calculation based on Latin hypercube sampling with distributed generators output as random variable.The programming model is solved by improved genetic algorithm with crossover and mutation operation probability and sequence self-adapting change.The test results of the IEEE-33 bus system verify the effectiveness of this method,and the efficiency of the improved genetic algorithm is proved by comparing with the traditional genetic algorithm.