引用本文:刘平,李国栋,杨金凤,刘兴高.集装箱装卸摆动最优控制快速数值求解算法[J].控制理论与应用,2019,36(8):1275~1282.[点击复制]
LIU Ping,LI Guo-dong,YANG Jin-feng,LIU Xing-gao.Fast optimal control numerical approach for the swing control of container load[J].Control Theory and Technology,2019,36(8):1275~1282.[点击复制]
集装箱装卸摆动最优控制快速数值求解算法
Fast optimal control numerical approach for the swing control of container load
摘要点击 2244  全文点击 977  投稿时间：2018-05-14  修订日期：2018-10-18
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2018.80352
  2019,36(8):1275-1282
中文关键词  集装箱装卸  最优控制  控制向量参数化  快速求解
英文关键词  container load  optimal control  control vector parameterization  fast calculation
基金项目  国家自然科学基金(61803060)，重庆市教委科学技术研究项目(KJQN201800635)
作者单位E-mail
刘平* 重庆邮电大学自动化学院 liuping_cqupt@cqupt.edu.cn 
李国栋 中国电子科技集团公司电子科学研究院  
杨金凤 重庆建筑工程职业学院  
刘兴高 浙江大学  
中文摘要
      为了实现起重机集装箱摆动最优控制,提出一种基于控制向量参数化（Control vector parameterization, CVP）方法的最优控制问题快速求解算法.首先,建立了以摆动能量最小为目标的集装箱装卸最优控制数学模型.其次,采用光滑化惩罚函数路径约束处理方法降低了模型求解难度.进一步,针对控制向量参数化方法微分方程组求解耗时长难题,结合网格划分提出改进四阶Runge-Kutta方法的快速CVP算法加快了最优控制问题求解速度.仿真测试针对不同位置的集装箱装卸任务进行.数值测试结果显示,相较于其他变步长求解方法,改进方法在得到相近求解精度解的同时,求解耗时明显减少,表明本文方法在集装箱装卸最优控制方面的应用价值.
英文摘要
      To achieve the optimal control for the container’s swing of crane, a fast calculation algorithm for optimal control problems based on Control Vector Parameterization (CVP) is proposed in this paper. Firstly, an optimal control mathematical model with minimum the swing energy of container load is developed. Next, the smoothing penalty method is employed to handle the path constraints so as to reduce the solving difficulty of the optimal control model. Furthermore, to tackle with the time-consuming issue of differential equations, an adaptive fast 4-th order Runge-Kutta method, which combines with grid subdivision, is proposed to accelerate the solving efficiency of optimal control problems. Numerical tests are carried out on container load tasks with different starting positions. Simulation results show that the proposed method is able to obtain similar solutions, while the computation time is obviously reduced when compared with variable step algorithms, revealing the application value for container load optimal control.