论文部分内容阅读
提出了一种使用神经元网络的非线性系统间接自校正预测控制器．这一控制方法是基于两个神经元网络，一个用于动态过程的建模，另一个用于获取控制信号．控制器的优化目标基于过程输出的向前多步预测．两个网络的在线学习均采用了非线性最小二乘法．
A nonlinear system indirect self-tuning predictive controller using neural networks is proposed. This control method is based on two neural networks, one for modeling dynamic processes and the other for acquiring control signals. The controller’s optimization goal is based on forward multi-step prediction of the process output. Both networks use non-linear least-squares for online learning.