摘要：多质量弹性扭转系统广泛存在于诸如机器人手臂及长轴距负载驱动等工业应用中。在对多质量弹性扭转系统结构分析和数学建模的基础上，利用Matlab辨识工具箱完成系统未知参数辨识，确定系统数学模型。以贝加莱PLC为控制单元、增量式编码器为速度传感器实现系统转速的闭环控制。为了消减弹簧传动和滤波延时引起的系统震荡，提出了灰色预测结合专家PID的控制策略。系统实验结果证明数学模型正确，控制性能良好。

关键词：多质量弹性扭转系统；建模；辨识；PLC；灰色预测；专家PID
Abstract: Multi-mass elastic torsion system widely exists in industrial application such as a robot arm and a long wheelbase load driver. After system’s structure analysis and mathematical modeling, Matlab toolbox was used to identify the unknown parameters. B&R PLC was used as the control unit and the incremental encoder was used as the speed sensor to realize the speed closed loop control. In order to reduce system oscillation which is caused by the spring driving and filtering delay, the control strategy that grey prediction combine with the expert PID was proposed. Experiment results show that the model is correct and the good control performance.

Key words: multi-mass elastic torsion system; modeling; parameter identification; PLC; grey prediction; expert PID