Arc Fault Identification of Wireless Communication Equipment Based on PSO-FNN

In the era when information transmission is highly dependent on radio communication, if the arc fault occurs in the power distribution system of its equipment, it will seriously affect the communication service and cause huge economic losses. In this paper, an arc fault identification method combining specific time-frequency analysis technology and machine learning algorithm is proposed. Firstly, from the frequency domain and time domain, the three features of the original current, VentropyHigh, VentropyLow, and Vkurtisis, respectively, were extracted to construct the feature vectors. Then, in order to solve the problem of model parameter setting, the particle swarm optimization algorithm (PSO) was combined with FNN for model training. After testing, the accuracy of the model on the training set, validation set and test set reached <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$96.6 \%, 96.8667 \%$</tex> and 96.6667 %, respectively, and the accuracy, recall and F1 value also performed well, with high prediction efficiency and real-time requirements.