摘要
采用IMC结构与人工神经网络方法解决了无法得到参考信号和系统非线性的问题,设计内模型神经网络控制器对噪声进行控制。通过仿真和实验证明,该控制结构能有效消除噪声中所含的周期噪声。非线性系统的仿真实例表明,内模型神经网络控制方法明显优于线性滤波X-LMS算法。
In order to solve the problem of getting the reference signal and the nonlinear system, the IMC structure and neural network are adopted. The IMC-neural network controller is designed to suppress the noise. Computer simulations and experiments show that this control structure can reduce tonal frequency noise effectively. A nonlinear example is given to support that the IMC-neural network control method is more efficient to the nonlinear noise control than the Filtered-x LMS algorithm.
出处
《电声技术》
2005年第11期58-60,共3页
Audio Engineering
基金
北京市教委资助科技项目(KM200311232137
KM200511232008)
北京市重点建设学科资助项目
关键词
有源噪声控制
IMC结构
人工神经网络
非线性系统
active noise control
Internal Model Control （IMC） structure
neural network
nonlinear system
