Power System Fault Diagnosis Method Based On Fuzzy Reasoning Spiking Neural P Systems

With the expansion of the power grid scale and the enhancement of interconnection degree,the power system fault will evolve into a large area of power outages and then cause significant economic losses to society if the fault can not be solved timely and effectively.When power system fault occurs,the dispatcher is required to solve the fault quickly and accurately,thus isolating the faulty area and providing the basis for the power supply restoration.And a large amount of fault multi-source information swarms into the dispatching center,which will not only provides diagnosis basis but also brings great difficulty to the dispatcher in diagnosis process.Therefore,it is of great practical significance to study an efficient fault diagnosis method for power system,which can be used to solve the fault diagnosis and ensure the safe and stable operation of the system.Fuzzy reasoning spiking neural P system is a computing model with the characteristic of intuitive graphical representation,parallelism,dynamic and uncertainty.It is generated by integrating fuzzy theory with membrane computing theory,which is used to solve uncertainty factors of practical problems.And the power system fault is a discrete dynamic evolution process with uncertainty factors,which consists of fault,protection action,circuit breaker trip and series of events.The characteristics of fuzzy reasoning spiking neural P systems make it be suitable for solving fault diagnosis problem.In recent years,it has been applied to power system fault diagnosis field and has got some developments.Therefore,this paper continues to study a fault diagnosis method based on fuzzy reasoning spiking neural P systems with the use of fault multi-source information obtained from the dispatching center.In this paper,the power system fault element identification method based on temporal fuzzy reasoning spiking neural P systems is given by using alarm temporal information.This method establishes the fault identification model firstly,and then uses the temporal consistency constraint to check the validity of the alarm information,thus correcting the initial confidence of the protection and circuit breaker action information and improving the identification result accuracy.Finally,this method is tested by examples of IEEE 39-node grid model and compared with the results of other diagnosis methods.A fault classification method of transmission line combined with wavelet transform and fuzzy reasoning spiking neural P system is given by using fault electrical quantity.Firstly,the wavelet transform and singular value decomposition are applied to extract the fault feature.Then the classification model based on fuzzy reasoning spiking neural P systems is used to identify the fault type on the basis of fault feature,thus determining the fault type.Taking a fault model of 500kV transmission line as an example,the fault classification method is verified under different fault initial angles,different fault resistances and different fault distances.The adaptability of the method in different line parameters and the presence of noise are analyzed.