A Study On Switch Cabinet State Detection And Fault Diagnosis System

With the rapid development of our economy, electric power facilities is closely related to people’s production and living, the stable operation of the electric power facilities or large equipment need to form a complete set of equipment to assist, in power distribution system, the need to be intelligent power distribution, electrical energy intelligence gathering and distribution switchgear plays a very important role, distribution switchgear can through to the system state detection and fault diagnosis to ensure the safe and stable operation of power distribution system. When a large equipment malfunction, rely on manual inspection or traditional detection methods for fault detection of various modules is a very difficult work, requires a lot of manpower and material resources, in this context, this article hopes to develop a have a certain intelligent switchgear to the system state detection and fault diagnosis of the switchgear intelligence by artificial neural network to realize.In this paper, the main research contents include the following:first is the state of the switch cabinet detection function, including the basic structure of the need to involve the switchgear, the basic principle of detecting state and state detection system hardware design and software design. Followed by fault diagnosis function of the switch cabinet, this paper designed a fault diagnosis system based on artificial neural network, through prior knowledge to training of neural network system, and has self learning ability makes fault diagnosis precision is increased. Then, this article has carried on the system the design of the communication module, and through CAN bus to realize the communication system. Finally this system, the simulation experiment is introduced the principle and method of experiment and the experimental results are analyzed in detail. Intelligent switchgear implementation based on artificial neural network in this paper can realize condition detection and fault diagnosis of equipment, the judgment of the trained system close to the expert judgment, an expert has only in one area often experience, and often also because forget to get the right judgment, and the expert system is designed in this paper can be all the expert’s experience as a priori knowledge on the system training, the limitations of the knowledge of the service, also overcome forgetting about the inherent nature of human, system through continuous learning, independent adjustment of reasoning rules, make the system similar to human’s independent thinking ability.In this paper, the system was tested, in terms of state detection, and carries out a comparative analysis on the actual measured value, in terms of fault diagnosis, through Matlab, the simulation experiment results show that this system can satisfy the use requirement. This system has the following advantages:this system can provide help for the safe and stable operation of distribution systems; After the system put into operation, the system can be run unattended, can save a lot of manpower and resources for the enterprise. In power grid failure, can through the various characteristics of fast fault diagnosis, mainly is the cause of the failure and locating fault location and provide decision support for troubleshooting, can save a lot of time for troubleshooting, also won’t because experienced older workers temporarily not to troubleshooting; Through to the equipment of all kinds of threshold setting, can make the system will failure can alarm in advance, to avoid the occurrence of failure; Can optimize system resources, balanced load, so as to achieve the goal of optimizing system. This system has a very wide range of application prospects, in office buildings, large business super, communities can be obtained in the application, has good market potential.