论文部分内容阅读
冷轧镀锌过程具有外界扰动多、时变大滞后、强过程非线性、生产过程中镀层厚度和生产线速度切换频繁等特点,控制难度大,目前多采用操作工手动模式进行控制,导致镀层厚度波动大、产品质量不稳定、切换过程缓慢等问题。针对上述问题,基于神经网络预测模型,采用“前馈+反馈”的控制器架构,研发并投运了镀锌厚度自动控制系统。实际运行效果表明,系统成功实现了冷轧镀锌过程的闭环自动控制,有效地减小了镀层厚度的控制方差与镀层厚度切换过渡时间。
Cold-rolled galvanizing process has the characteristics of more disturbance, longer time-lag, strong process nonlinearity, frequent change of plating thickness and production line speed in the production process and so on, which is difficult to control. At present, the manual mode of operator is mostly used to control the thickness of the coating Volatile, unstable product quality, slow switching and other issues. In view of the above problems, based on neural network prediction model, “feedforward + feedback ” controller architecture was used to develop and put into operation a galvanized thickness automatic control system. The actual operating results show that the system successfully closed loop automatic control of cold-rolled galvanizing process, effectively reducing the control variance of coating thickness and switching the transition time of coating thickness.