激光水下目标探测中混沌背景信号重构的研究

对激光水下目标探测中混沌背景信号的重构问题进行了研究．讨论了混沌时间序列的动态特性，并实际计算了激光水下目标探测中混沌背景信号的时延、混沌维数等有关特征参量．在阐述神经网络重构时间序列模型机理的基础上，提出用神经网络局部预测法重构水下目标探测中混沌背景信号，最后在成功地重构出混沌背景信号的条件下，利用预测误差检测到水下目标探测中的有用弱信号．实验结果表明这种方法是比较有效的．

On the Reconstruction of Chaotic Background Signals Received during Underwater Target Detection with Laser

The reconstruction of chaotic background signals received during underwater target detection with laser is specifically studied. The dynamic characteristics of chaotic time series are discussed and the relevant characteristic parameters such as the signal time delay and the chaotic dimension are calculated. Based on the mechanism of the model for the reconstruction of time series with the neural network, the reconstruction of chaotic background signals during underwater target detection with local prediction method by the neural network is proposed. In the case where the signals have been successfully reconstructed, useful weak signals are detected with prediction errors. Experimental results show that the method proposed is effective.