Evaluate And Research On Reliability Of Distribution System

The power distribution system, directly connected with the user which was the endof the whole power system, has direct influence upon the quality and reliability of thepower supply, therefore, how to evaluate the reliability of the power distribution systemaccurately and rapidly has become the research hot spot at present. This paper combinedthe power distribution system reliability evaluation with blind number theory,successful probability method, frequency duration method and related theory, especiallyresearched the transmission line which is the important link in power distributionsystem. The research achievements are as follows.First, this paper introduces blind number theory in allusion to the credibilityproblem of original data which exists in distribution network reliability evaluation, andintroduced a new reliability assessment method which based on blind number theoryand the probability of success, we take IEEE-RBTS BUS6bus bar4as an example,and verify the method, the results of the calculation show that the proposed algorithm iscorrect and feasibleSecond, we molded the distribution network reliability which contains DGapplying a frequency duration method.. Using the markov chain researched the systemstates transfer relationship and transition frequency between states, established thepower distribution system state transfer diagram. Using the frequency, cumulativefrequency and shift frequency derived the load point index formula and using a simpledistribution network validates the proposed method.Third, this paper proposes a genetic neural network to evaluate the reliability of thetransmission line which is the important part of the distribution network. Using geneticalgorithm optimizes the threshold and initial weight on BP network, the convergencespeed of optimized network became quickly, the global optimization ability becamestronger and the evaluation results became more objective and accurate.