Application Of Spiking Neural P Systems To Power System Diagnosis

With the rapid development of chinese economics,the demand of electricity is increasing sharply.Meanwhile,the stable running of the power system has become very important to the nation's economy and people's livelihood.However,it's difficult to avoid failure due to the power system has large scale,complex structure,and exposed to the harsh natural environment in a long time.At the same time,it's great significance to the stable running of the subway that the subway traction power supply system,as the subway energy system,operates safety and reliably.However,in recent years,because of the faults of subway traction power system,the accidents of the train outage and late have happened sometimes.Therefore,what we can do is diagnosing and isolating the faults quickly.However,in the actual running,the traditional method of fault does not solve the problem of power system diagnosis well,misjudgment and miscarriage of justice still occur sometimes.Hence,in order to solve the fault diagnosis problem better,we should not only improve the original methods,but also explore new fault diagnosis methods.Spiking neural P systems is a class of distribution and parallel computing model with good dynamics.Based on this,many studies apply it to solving practical problems.In this paper,the spiking neural P systems is applied to solve the power system diagnosis and subway traction power system diagnosis,including the following three points:(1)Giving the waveform similarity transmission fault reliability.In this paper,we use the wavelet transform theory to analysis the change of amplitude and harmonic of the waveform signal when the transmission line have faulted,and the waveform correlation coefficient is used to reflect the change of the amplitude and harmonic of the line fault.In order to verify the waveform similarity transmission fault reliability can reflect the transmission line fault,this paper used PSCAD building model and simulation 180 di:fferent faults to verify its effectiveness;(2)Applicating the spiking neural P systems into the fault phase selection.Building the fault phase selection diagnosis model based on spike neural P systems,this paper used 6 kinds of eigenvalues as phase selection evidence and give the calculation methods respectively.Giving the fault phase selection reasoning algorithm which based on fault phase selection diagnosis model and using PSCAD build a model and simulation 450 different fault types to verify the method effectiveness;(3)Using the spiking neural P systems into the fault diagnosis of subway traction power system.Based on the network topology analysis method,this paper gave the fault area determination method to determine the suspected fault components.Building fault diagnosis models based on spike neural P systems for every suspected fault components.According to the fault diagnosis models,we can acknowledge the fault reliability of suspected fault component by calculating,respectively.Thus,we can acknowledge the fault component.