| An overhead power line which is main way of transmission of electrical energy is the lifeblood of transmission system. And its running status is directly related to the entire power system security and reliability. Once the transmission line problems will cause huge economic losses. So the fault diagnosis of the overhead power line has a significant economic and practical significance.With the strong development of robot technology, fault diagnosis method for robotic inspection data also gradually on the rise. Because the Existing fault diagnosis methods mostly for manual inspection results or online sensor data, whose scope of application was narrow, the existing fault diagnosis methods difficult to apply for robot inspection data. The main issues are as follow:(1)There is no uniform standards of fault diagnosis.(2)It’s very difficult to get fault data,and lack of simulation data pre-simulation.(3)There have no fault diagnosis program for the whole overhead power lines.In response to these problems, the paper carried out the work as follow.Firstly, combined with the national standards, this paper summarized the fault detection methods of overhead power, classified for fault detection method, and described the techniques principles of each type of fault detection methods in detail. Secondly,the paper proposed a new fault simulation data generation algorithm, we can generate a lot of complete, consistent, and equipment failure data.Thirdly,this paper studied the basic Bayesian network overhead power line fault diagnosis method based on Bayesian network, and a precise formula for derivation of modeling, parameter learning and seasoning mechanisms were given by the paper, and experiments show that the model is an accurate and efficient troubleshooting tool. Finally, I developed a overhead power line fault diagnosis system, the detailed implementation of the feasibility analysis were given and finished the basic functions of the system. |
PDF Full Text Request