用解析模型描述层流冷却过程的热轧带钢卷取温度控制系统在换规格时自适应能力较差 ,因为系统的自适应系数体现的是均值效应 ,对于动态随机不确定复杂过程的描述与控制往往无能为力·针对这一不足 ,采用小脑模型连接控制 (CMAC)网络进行动态随机自适应控制 ,取得了令人满意的结果·模拟冷却结果表明 ,小脑模型连接控制网络充分发挥了模糊联想记忆的功能 ,可以对没有精确数学模型的不确定复杂过程进行精确控制 ,增强了系统自适应能力 ,它是性能优良的前馈网络· The analytical model describes the laminar cooling process of the hot-rolled strip coiling temperature control system in changing specifications adaptive ability is poor, because the system’s adaptive coefficient reflects the mean effect, the dynamic random uncertainty of the complex process description and Control is often powerless. In response to this problem, the CMAC network is used to control the dynamic stochastic adaptive control, and satisfactory results have been obtained. The simulated cooling results show that the cerebellar model connected control network fully exerts the effect of fuzzy associative memory Function, which can accurately control the uncertain complicated process without accurate mathematical model and enhance the system adaptive ability, which is a good feedforward network ·