Research On Fault Section Location Methods For Distribution Networks With Distributed Generation

With the transition of global energy structure to clean,low-carbon mode and people's increasing concern for environmental protection,distributed generation technology with the advantages of flexible power supply and environmental protection has been widely applied.However,due to the massive access and high penetration of distributed generation,traditional distribution networks with unidirectional power flow are transformed into active distribution networks with bidirectional power flow,which also brings challenges to fault section location.Studying fault section location of distribution networks with distributed generation is of great significance to improve the power supply reliability and promote the development of distributed generation technology.To solve the problem of misjudgments of fault section location models based on intelligent algorithms,a generalized integer linear programming model of fault section location for distribution networks is put forward.The constraint conditions and a one-to-one mapping objective function are built up according to the upstream and downstream location relationships between terminal devices and line sections.The fault line section can be found out quickly and accurately by algebraic operation instead of logical operation.The model is not only applicable to distribution networks with single power supply,but also available to distribution networks with distributed generation,which can accurately locate single fault occurring on the main feeder or laterals as well as multiple faults occurring simultaneously on different laterals.As for information distortion,a fault tolerance method based on power access paths is proposed.Firstly,the feedback fault information of terminal devices is processed,the distorted information is found and corrected.Then,the corrected information is put into the integer linear programming model to determine the real fault section.To solve the problems that fault section location method based on matrix algorithm has complex judgement criteria and weak fault-tolerant capacity,a method based on reachability matrix for fault section location in distribution networks with distributed generation is put forward.Different from the idea of using adjacency matrix to describe the network topology in traditional matrix algorithm,the reachability matrix which contains more information and has stronger global observability is used for topology description and matrix operation.Simple and intuitive fault judgement criteria are set up to locate accurately single fault occurring on the main feeder or laterals as well as multiple faults occurring simultaneously on different laterals.To deal with information distortion,a fault tolerance method based on Bayes' theorem is proposed.Fault tolerance process is applied on the results obtained from fault section location method based on reachability matrix,conditional probability is used to eliminate the line sections with rather small possibility of occurring a fault to determine the real fault section.Finally,the effectivenesses of proposed fault section location methods and fault tolerance methods are validated by test cases.