Risk Assessment Of Distribution Network Operation Considering Line Fault And Photovoltaic Output Randomness

Nowadays,due to the increasingly complex grid structure of the distribution network,the increasing penetration rate of renewable energy,and the increasing proportion of electric vehicles,the uncertainty in the distribution network is increasing,which makes the distribution network.The safe operation is facing enormous challenges.Risk assessment is an important means to quantify uncertainties.The application of risk assessment technology in dispatching operation of distribution network can scientifically and reasonably quantify the adverse effects of various uncertainties such as equipment failures and distributed power output on the operation of the distribution network.Therefore,this paper focuses on the risk assessment of the fault uncertainty of the equipment in the distribution network and the uncertainty of the output of the distributed power supply.Firstly,this paper studies the prediction method of time-varying failure probability of feeder network feeder.Based on the Fokker-Planck equation,the time-varying fault probability prediction of feeders under different operating scenarios is proposed,and the Fokker-Planck Equation(Fokker-Planck Equation)is predicted by Support Vector Machine(SVM).The parameters in FPE avoid errors caused by subjective experience.Based on the correlation feature optimal feature subset evaluation algorithm,the optimal prediction feature variables are filtered out,which makes the SVM prediction model more accurate.In addition,considering the imbalance between the fault sample and the normal sample will bring errors to the data mining model.This paper uses Synthetic Minority Oversampling Technology(SMOTE)to reduce the imbalance between samples.This method can effectively improve the accuracy of the prediction.Finally,the method proposed in this paper is verified by taking the historical data of distribution network in a certain area as an example.Secondly,after obtaining the time-varying failure probability of the distribution line,the non-sequential monte carlo simulation method is used to calculate various risk indicators.Compared with the enumeration method risk assessment method,the non-sequential monte carlo method is easy to consider the change of the topology structure of the distribution network,and the simulation of the random failure occurring in the distribution system is closer to reality.Through the analysis of the example,the risk of loss of load can be calculated and the decision-making reference can be provided for the dispatching operators.Then,for the integrated of high-density distributed photovoltaics in the distribution network,there is operational risk in the operation of the distribution network.This paper first analyzes the main factors affecting photovoltaic power generation,and uses the Dynamic Bayesian Network(DBN)method to predict the probability distribution of photovoltaic power output.Finally,the enumeration method,monte carlo method and cumulant method are used to calculate the voltage rise and branch power risk caused by the power injection of photovoltaic power generation.The actual distribution network of a city and the IEEE-33 bus test system are selected as test systems to verify the correctness of the method proposed in this chapter.