Distirbution Network Fault Interval Location Based On Intelligent Algorithm

Distribution network fault location is the premise and ensurance of realizingdistribution network automation, which plays a key role in the isolation ofdistribution network fault and shortening the outage time.At the same time theinformation from the feeder terminal unit is apt to be distorted and lack whenuploaded, so it needs high fault tolerance and rapidity when fault location in choosingmethods, based on which this paper is devoted to the study of application of improvedelectromagnetic algorithm and distribution estimation algorithm in distributionnetwork fault interval location in rapidity and fault tolerance.In this paper two kinds of nodes in distribution network fault location-intelligentautomation on the spot and centralized regulation feeder automation will besummarized firstly, especially for the last one,in which the advantages anddisadvantages are discribed for various methods.The optimization effect of the electromagnetic algorithm in discrete domain isgood, in this paper the electromagnetic algorithm is put in the fault location ofdistribution network,but the problem of parameters sensitive exists in the process,which effects the positioning,by introducing the adaptive step length in local search,eliminating distance factor in force calculation formula,and accelerating formulaparticle migration steps, the accuracy and rapidity of the fault location will beimproved. In the same time, the GA search-crossover operator is brought in to solvethe problem of falling into local optimum.On the basis of which the modulesdividision thoughts was brought in to build the mathematical model of the distributionnetwork,which effectively reduced the dimension of mathematical model of thedistribution network. The simulation results showed that compared with the standardelectromagnetic algorithm the improved electromagnetic algorithm only needed quitea small scale of population and iterations, the fault location accuracy also hascorresponding improvement.Different from the traditional heuristic algorithm in dealing with the optimationproblem in view of microstructure mathematical modeling, the distribution estimationalgorithm is focused on biological space overall on macroscopic analysis, which usesa new way of evolution.In this paper the distribution estimation algorithm isapplicated in the distribution network fault location, the fault section vector isregarded as optimal solution, by establishing a uniform probability model and building the more concise sampling method, the probability of the correct solution inthe solution space will gradually improved and the fault location of distributionnetwork will be achieved.The simulation results showed that the application of thedistribution estimation alogrithm in the distribution network fault location had fasterfault location speed and good fault tolerance.