论文部分内容阅读
研究了利用前向神经网络对混沌光学系统进行混沌加速系统辨识的可能性。计算机数值仿真发现，利用三层前向神经网络混沌光学系统辨识器，在基于混沌动力学角度的修正ＢＰ算法（混沌加速ＢＰ算法）支持下可克服由常规ＢＰ算法导致的辨识时间长的缺点，在较少的训练次数内即可对布拉格声光双稳混沌系统进行良好的系统辨识。此研究结果表明，在混沌加速ＢＰ算法的支持下，三层前向神经网络可用来快速处理混沌光学时间序列以进行相应的动力学重构。
The possibility of chaotic system identification using chaos optics is studied by using the feedforward neural network. Computer numerical simulation shows that the chaotic optical system recognizer based on three-layer feedforward neural network can overcome the shortcoming of long recognition time caused by the conventional BP algorithm with the support of the modified BP algorithm (chaotic acceleration BP algorithm) based on the chaotic dynamics. The Bragg acousto-optic bistable chaotic system can be well recognized systematically with less training times. The results of this study show that the three-layer feedforward neural network can be used to quickly process the chaotic optical time series for the corresponding kinetic reconstruction supported by the chaos acceleration BP algorithm.