Research On The Prediction Of Failure And Outage Of City Distribution Network Based On Reliability Index

The distribution network is directly connected to power users,and the reliability of the power supply of the distribution network is closely related to the production and life of users.Fault outage is an important factor that affects the reliability of distribution network power supply.The prediction of distribution network fault outage has practical significance and theoretical value to improve the power supply reliability of distribution network.In this thesis,the fault outage prediction models of distribution network based on IFA-RVM and SA-PSO-RVM are proposed to study the fault outage index of distribution network.There are many factors and uncertainties influencing power outages in the distribution network,which leads to the difficult problem of forecasting.In this thesis,the analytic hierarchy process(AHP)is used to classify the influence factors of distribution network fault outage,calculates the influence weights of the factors of each layer relative to fault outage,and ranks the results of the weights.The evaluation system of influence factors of fault outage is put forward by selecting the six influencing factors which are in front of the order.For the standard firefly algorithm(FA)in the process of optimizing the kernel function parameters of the RVM,the convergence speed becomes slower and the accuracy decreases.The adaptive inertia weight is introduced to optimize the firefly position update formula.At the same time,the search step factor is dynamically adjusted to balance the global and local search capabilities of the algorithm,resulting in an improved firefly algorithm(IFA).The experimental result shows that the optimization speed and accuracy of the IFA algorithm are improved compared with the standard FA algorithm.Aiming at the problem of particle swarm optimization(PSO)algorithm in the process of optimizing the parameters of the RVM kernel function,it is easy to fall into the local optimal value.The probabilistic jump mechanism of the simulated annealing algorithm(SA)is introduced into the PSO algorithm,and the simulated annealing-particle swarm algorithm(SA-PSO)is proposed.The experimental result shows that the SA-PSO algorithm has better optimization performance than the PSO algorithm,and it can jump out of the local extreme to get the global optimal solution.Using the data of a certain power grid from 2017 to 2019,the two prediction models proposed in the thesis are used for experimental verification and comparative analysis.The result shows that for the short-term prediction,the correlation coefficient between the predicted value and the actual value of the SA-PSO-RVM model is 0.0029 higher than that of the IFA-RVM model.For long-term prediction,the correlation coefficient between the predicted value and the actual value of IFA-RVM model is 0.0022 higher than that of SA-PSO-RVM model.