Study on Anomaly Detection of Power Communication Signal Based on Fuzzy Neural Network
Because the existing anomaly detection methods have low detection accuracy and can not effectively detect abnormal signals,this paper proposes an anomaly detection method for electric power communication signals based on fuzzy neural network.The collected sample data are input into the network,and stable network parameters are obtained through unsupervised learning and training,and important feature information is learned and extracted.The known data characteristic indexes are coded,and the abnormal signal characteristics are extracted by demodulation,and input into the fuzzy neural network.Combined with particle swarm optimization,the network is trained to obtain the global optimal value.Fast Fourier Transform(FFT)is used to transform the training data into signal amplitude,and to judge whether the signal is abnormal,so as to complete the signal abnormality detection.The experimental results show that the accuracy of 10 groups using the proposed method is above 99%,and the results are in line with expectations,and the anomaly detection effect is good.
fuzzy neural networkelectric power communicationsignal abnormality detection
NETL
NSTL
万方数据
