论文部分内容阅读
论述了小波神经网络用于信号分类识别的模型结构，在此基础上，充分利用小波变换时频分析的局部化特性，提出了一种改进的网络结构，建立了非显式小波的网络的学习算法。计算机模拟表明，该结构提高了信号分类识别的精度和灵敏度。
This paper discusses the model structure of wavelet neural network used for signal classification and recognition. Based on this, taking full advantage of the local characteristics of time-frequency analysis of wavelet transform, an improved network structure is proposed and the learning of non-explicit wavelet network is established algorithm. Computer simulation shows that the structure improves the accuracy and sensitivity of signal classification and recognition.