Neural networks data fusion algorithm of electronic equipment fault diagnosis

In order to judge accurately the fault component of an analog circuit, a fuzzy neural networks fault classifier was designed based on BP neural networks and fuzzy logical theory, and it was used in the photovoltaic radar electronic equipment fault diagnosis. By measuring the temperature and voltage of the circuit component, the membership function of the two sensors to the circuit component was obtained, the data fusion was done by using a fuzzy BP neural networks classifier, the fusion fault membership function of all the fault-doubted circuit components was obtained, and the real fault component was found based on the fusion data. By comparing the diagnosis results based on a separate original data and fused date respectively, it was shown that the latter is more accurate than the former in circuit fault recognition.