论文部分内容阅读
针对神经网络对系统辨识时，在一次采样周期内只学习一次样本对的低效率情况，采用一种改进措施，即在一次采样周期内进行多个样本对的学习，同时还采用离线辨识和在线辨识结合的方法，使神经网络能快速、准确地辨识系统的特性．
When the neural network identifies the system, only one sample pair’s inefficiency is learned in one sampling period. An improvement measure is taken to learn multiple sample pairs in one sampling period. Offline identification and online Identify the combination of methods, so that neural network can quickly and accurately identify the characteristics of the system.