论文部分内容阅读
研究了利用ＷａｖｅＡＲＸ神经网络构造非线性动态过程的非参数化逼近模型，在分析其逼近偏差的基础上提出了一种非线性次优滤波器设计方法，用于非线性动态过程的故障检测。理论分析和仿真研究证实了其可行性和有效性。
A nonparametric approximation model for constructing nonlinear dynamic process using WaveARX neural network is studied. Based on the analysis of the approximation bias, a nonlinear suboptimal filter design method is proposed for the fault detection of nonlinear dynamic process. Theoretical analysis and simulation studies confirm its feasibility and effectiveness.