Research On Multi-level Distribution Network Fault Diagnosis Method Based On The Multi-Source Information

With the continuous development of modern technology and the improvement of people’s living standards in our country, electricity has become an indispensable part of our life. Distribution network is an important link in power system, and its security and reliability is the key to ensure power system to operate stably. Once the failures happen or operating condition is threatened by outside factors in distribution system, the failures may develop into cascading failures if there are no prompt measures being taken, even bring huge economic losses as well as inconvenience brought about to consumers. However, fault diagnosis is the premise of the processing, the current fault diagnosis methods of distribution network most merely on switch and protection action information, but when situation that the switch or protection refuses to action or makes false action or complex fault happens, the traditional method that only relays on the single data resource has not been able to diagnose the fault components accurately. Therefore, research on fault diagnosis methods which is able to make the most use of multi-source information are of magnificent significance. By analyzing the current development and existing problems in fault diagnosis and source of fault information of power system, a dynamic hierarchical fault diagnosis method based on multi-source information is proposed in this thesis.First of all, from the perspective of difficulty and amount of obtaining the failure data, this paper put forward a dynamic hierarchical fault diagnosis method based on multi-source information, which including switch layer used for diagnosis of simple fault, feeder layer strived to resolve complex fault in the case of abnormal switch and protection information, and substation layer to judge multi-type fault in the complex system. Simultaneously, it adopted dynamic diagnosis strategy and adjusted diagnostic entrance and structure longitudinally according to the fault characteristics. This method enhances the adaptability of each layer diagnosis, improves the efficiency and accuracy of fault diagnosis;Secondly, from the perspective of equipment and construction cost of the actual power system, this thesis proposes hierarchical fault diagnosis approach for smart power grid with information fusion of multi-data resources. The method includes switch layer that can determine primary fault candidate set and multi-source information fusion diagnosis layer based on directional weighted fuzzy petri net. The method can automatically adapt to the network topology changes, and has good versatility and fault tolerance;Finally, considering identifying distribution network fault type accurately is the important guarantee of rapid and accurate processing faults, an identification method of distribution network based on immune neural network is presented after comprehensive analyzing the fault feature. The method uses wavelet transform to extract the fault feature seen as the immune neural network’s input, then adopts immune algorithm to optimize the weights of neural network to avoid the neural network algorithm falling into local minimum. From the simulation we can see that the method can accurately realize identification for fault type in any kind of fault mode, and has good robustness and convergence.