【摘要】 目的　用神经网络实现 Adaline自适应滤波权值调节 .方法　在 Adaline自适应滤波器中 ,滤波器的权值调节是通过 LMS自适应算法进行调整的 ,但 L MS的自适应算法的收敛速度慢 ,从而大大影响了 Adaline自适应滤波器的滤波性能 ,作者利用 TH网络进行TH- Adaline滤波器系统辩识 .并给出了 Adaline滤波器权值输出变化曲线的仿真结果 .结果与结论　利用 TH神经网络来实现滤波器权值调节的 Adaline自适应滤波器运算速度比运用 LMS算法的 Adaline自适应滤波器的收敛速度快更多还原

【Abstract】 Aim\ To realize a self-adaptive filter with artificia l neural network. Methods\ LSM self-adaptive algorithm has a big problem in the con stringency speed which results in the bad characteristic of the Adaline filter. A TH neural network method is presented in this paper. The simulation result is given too. Results and Conclusion\ The learning rate of this kinds filter is higher than the Adaline self-adaptive filter that uses LMS algorithm.更多还原