迭代学习控制针对具有重复运行性质的系统 ,利用系统实际输出与期望输出的偏差信号 ,产生新的控制信号 ,使得系统跟踪调节性能得以提高。根据张减过程轧制前后钢管壁厚的实测数据和钢管的特征数据 ,采用迭代自学习控制算法 ,提出了无缝钢管张减过程的平均壁厚控制的迭代自学习控制。该控制技术在轧制过程中在线自适应调整各轧制机架的稳态转速分布 ,补偿由物理参数的时变不确定性和建模误差造成的轧辊转速分布参数误差。计算机仿真和实际应用的结果表明该学习控制技术的有效性 Iterative learning control for systems with repetitive properties, the use of the system output and the desired output of the actual deviation of the signal, resulting in a new control signal, making the system tracking adjustment performance can be improved. According to the measured data of the pipe wall thickness before and after rolling and the characteristic data of the steel pipe, the iterative self-learning control algorithm is proposed and the iterative self-learning control of the average wall thickness control during the process of reducing the seamless steel pipe is proposed. The control technique adjusts the steady state speed distribution of each rolling stand adaptively during the rolling process and compensates the parameter error of the roll speed distribution caused by the time-varying uncertainty of the physical parameters and the modeling error. Computer simulation and practical application of the results show that the learning control technology is effective