The Research of Building Electrical Fault Diagnosis Based on Experimental Platform

In the building electrical system, the faults need to be classified and located accurately. In this essay, the building electrical test platform was utilized as the research and verification object of fault diagnosis to supply the test data and verify the availability of different algorithms. Choosing the neural network to verify the model in the fault diagnosis system, the result indicated that RBF Network and BP Network can both diagnose the fault and the rate of convergence is satisfactory, and the RBF Network has the advantages of the accuracy and stability in this experiment. The application data collection system can collect the test data, and this function realizes the experimental platform's data collection, storage and waveform display, smart fault diagnosis, etc.