Researches On Fault Identification And Fault Location Of Power System Based On Spiking Neural P Systems

Fault identification and fault location is always an important research content of power system fault diagnosis.When transmission lines fault in power system,a mass of alarm information which may also be missing or distorted will produce in automatic devices of power system.If alarm information can't be accurately and timely processed and fault location can't be identified by operator,a large area of power failure may happen.Thus,processing of alarm information is particularly important in fault recognition.Meanwhile,when distributed generations are introduced into distribution network,power flow structure and operation mode of distribution network will change.Also the traditional protection strategy is no longer applicable and there exist problems of inaccurate positioning and complex process of positioning.Spiking neural P systems which have the characteristics of powerful distributed parallel computing,information processing and graphic image are suitable to solve problems of redundant fault information,complex diagnosis process,misjudgment in fault recognition and fault location.Therefore,based on spiking neural P systems,fault identification of transmission network with voltage level above 220 k V and fault location of distribution network with distributed generation have studied in this paper.In terms of fault identification,this paper introduces the triangular fuzzy number based on Fuzzy Reasoning Spiking Neural P Systems(FRSNPS),redefines parameters such as neurons,synapses,spikes,firing rules and firing conditions,proposes Fuzzy Spiking Neural P Systems with Triangular(TFSNPS)and compares it with FRSNPS.Then,220 k V and its above ring network are studied,combined with the basic principle of longitudinal differential protection,fault diagnosis models based on TFSNPS are constructed and its reasoning algorithm is proposed.Finally,examples of single fault and incomplete information fault are illustrated to analyze and verify the model and its algorithm.The method of fault identification is proposed in this paper can accurately identify fault of power system transmission network,has good fault tolerance and is applied in the uncertain and incomplete information circumstances.In terms of fault location,this paper redefines parameters such as neurons,synapses,spikes,firing rules and firing conditions based on Spiking Neural P Systems with Anti-Spikes(ASNPS),proposes Improved Spiking Neural P System with Anti-Spikes(IASNPS)and compares it with ASNPS.Then,distribution network with distributed generations(DG)are studied,combined with the basic principle of fault location,fault location models based on IASNPS are constructed and its reasoning algorithm is proposed.Finally,two cases of simple multi-power distribution networks and distribution network with DG are analyzed and verified the model and its algorithm.Meanwhile,single fault,multiple faults,incomplete fault information and fault information distortion are included in each case.The results show that the proposed fault location method can accurately locate single fault,multiple faults and faults that contain uncertain and incomplete fault information in distribution network with DG.