The Parameter Detection of Weak Signal Based on Chaos and Neural Network

This paper dealt with the parameters detection of weak signal, which based on chaos and neural network. According to the characteristics of chaotic time series, chosen Elman network as neural network, constructed the network detection model though solving the correlation dimension of chaotic time series to determine input and output dimensions of the network, adopted single-step prediction method to detect the weak signals directly from the chaotic background under the chaotic state. This method breakthrough the traditional chaos detection principle, can detect the time-domain parameters of weak signal, and has advantages of wide measuring range ,high precision in approximating target, and embed in the digital oscilloscope easily. The experimental results show that this method is of high practical value.