Power System Fault Diagnosis Methods And Applications Employing Information From Multiple Sources

Prompt fault diagnosis can help restore the power system concerned quickly,and thus reduce outage costs.Meteorological factors,as well as other impacting factors,represent some of the main causes of power system faults,and how to appropriately take into account of the influences of these impacting factors in power system fault diagnosis is a challenge task.Given this background,the influences of meteorological and other impacting factors are systematically investigated,and based on an existing fault diagnosis method the meteorological and other impacting factors at the fault occurrence period are employed to analyze the fault cause.First,the outage area is identified so as to attain the suspected fault sections using a network topology analysis method.Then,a simplified fault diagnosis model is presented to identify the fault section(s)from suspected ones by utilizing the state information of protective relays and circuit breakers.Finally,data from the concerned contamination monitoring system,lightning monitoring system,weather pre-warning system,and monitoring equipments for other external environmental factors,together with the estimated fault section(s),are used for analyzing those factors having impacts on fault occurrence,so as to assist the operators to locate and clear the fault(s),and ultimately speed the system restoration procedure.The traditional backup protection has some drawbacks,including difficult protection setting,long time delay and complicated coordination.Given this background,a fault section identification and wide-area backup protection strategy is proposed based on limited phasor measurement units(PMUs).First,the whole power network is divided into several backup protection zones(BPZs)based on the placement of PMUs.Once a fault occurs,the electrical quantities are utilized to identify the faulted BPZ,so as to quickly narrow down the scope of suspected fault sections.Then,an analytic model based fault section identification method and a wide-area backup protection strategy are presented using alarm information of main and backup protective relays,power direction relays,and circuit breakers.The proposed strategy will be activated only if the main protection relay fails to clear the fault.The well-established Tabu Search(TS)algorithm is utilized to seek the optimal fault hypothesis,and determine the malfunctioned protective relays and/or circuit breakers.Based on the identified malfunctions,the corresponding backup protection scheme is initiated to prevent the fault from spreading.Finally,several test cases are carried out on the IEEE 39-bus power system,and it is demonstrated that the proposed method is feasible and fault-tolerated.Aiming at the problem of large-scale increasement of distributed power generation,electric vehicles and controllable load in Chinese distribution network,this thesis points out the shortcomings of protection device and fault diagnosis method in traditional distribution network,and presents an intelligent fault diagnosis framework based on micro synchronous phasor measurement units(?PMUs).Using the high-precision electrical data information collected by?PMUs and the node over-current information monitored by feeder terminal units(FTUs),this thesis propose an intelligent fault diagnosis and localization method for active distribution networks.After a fault occurs,the area of suspected faulty devices is determined based on the power failure area.Secondly,the fault diagnosis model of power distribution system based on multi-source information is established by using the information of ?MUs and FTUs collected by the control station in the distribution system.Secondly,combining the alarm information from ?PMUs and FTUs received by the control master station,the fault diagnosis method.of the distribution system based on analytical model is established.Then,the optimization algorithm is used to solve the optimal solution and find the faulty component.Finally,the fault diagnosis results are evaluated.The last part of this thesis summarizes the whole text,and makes a prospect for the future research in this thesis.