Substation electric power equipment detection based on patrol robots

The expansion of power grid scale not only increases the transmission capacity, but also increases the probability of power plant facilities failure. The large scale of power grid and its high voltage make fault detection have heavy workload and high risk. In this paper, patrol robot, infrared imaging technology for detecting equipment faults and support vector machine (SVM) for identifying infrared image of faulty equipment were briefly introduced. Then, SVM for identifying infrared image of faulty equipment was simulated and analyzed with MATLAB software and compared with information entropy method. Then, patrol robot which applied two recognition methods in substation of X city power supply bureau were operated for one month. The results showed that the recognition accuracy of SVM was above 97% in the simulation experiment, which was significantly higher than that of information entropy method. In actual operation, the efficiency of patrol robot was higher than that of the traditional manual patrol, and the failure recognition rate of patrol robot which applied the two methods was close to the simulation results.