基于优化遗传小波网络的混沌时间序列预测

研究利用小波神经网络(WNN)预测混沌时间序列。提出了一种改进的小波神经网络训练算法，该方法融合了遗传算法和梯度下降算法两种方法，在遗传算法中嵌入梯度下降算法以解决遗传算法不具有的细节搜索能力，对遗传算法训练后的小波网络再次利用梯度下降算法寻找最优点。对Henon映射混沌时间序列的预测证明了该方法的有效性，实验结果表明该算法能确保小波网络收敛和具有较高的预测精度。

Forecasting chaotic time series based on improved genetic WNN

The chaotic time series forecast using Wavelet Neural Networks (WNN) was researched in this paper. An improved training method for WNN was presented. This method combines the Genetic Arithmetic (GA) and gradient descent BP method, and the BP method was embedded in the GA operation in order to resolve the GA's limitation in detail search capability. In the last step of this method the WNN trained by GA searches the best solution using BP method once again. The experiment on predicting the chaotic time series from Henon map validates the performance of the method in this paper; the experimental result also shows the method could assure the WNN convergence and have high forecasting precision.