基于前摄—反应的泊位岸桥鲁棒性集成调度  

Robust integrated scheduling for berth and quay crane based on proactive-reactive method

作　　者：张振毓[1] 陆志强[1] 韩笑乐[1] 

机构地区：[1]同济大学机械与能源工程学院,上海201804

出　　处：《计算机集成制造系统》2014年第5期1181-1190,共10页Computer Integrated Manufacturing Systems

基　　金：国家自然科学基金资助项目(70771065;71171130);2013智能制造装备发展专项~~

摘　　要：针对不确定因素下集装箱码头岸边侧泊位岸桥资源的协同调度问题,提出前摄调度计划和与之匹配的反应策略相结合的决策框架,并在前摄调度计划中引入了时间缓冲,以吸收船舶到港时间不确定带来的影响。以质量鲁棒性和解鲁棒性为目标函数,建立了不确定环境下连续型泊位岸桥集成调度的数学模型,并构建了具有双层决策结构的优化算法对模型进行求解。其中,上层采用非支配排序遗传算法框架搜索关键决策变量编码,下层运用带缓冲的前摄启发式算法对上层解码,并结合Monto Caro抽样及所设计的两种反应策略进行仿真评估。通过数值实验验证了所提方法框架与算法的有效性。Aiming at the collaborative allocation for berth and quay crane resources of container terminals under uncertain condition,a decision framework by combining proactive scheduling plan with corresponding reactive policy was proposed,and the time buffer was introduced into proactive scheduling plan to avoid the influence of the uncertain vessel arrival time.An integrated mathematical model for continuous berth and quay crane was formulated with the objective of balancing both quality robustness and solution robustness,and an optimization algorithm of bilevel decision structure was built to solve this model.In this algorithm,NSGA Ⅱ framework was applied in upper level to search key decision variable codes,and a proactive heuristic with time buffer was applied in lower level to decode each individual from upper level,whose performance was evaluated by simulation under Monte Carlo sampling and two designed reactive policies.Numeral experiment was used to show the effectiveness of proposed method.

关 键 词：集装箱码头 前摄—反应 不确定性 集成调度 MONTE Carlo抽样 

分 类 号：TP29[自动化与计算机技术—检测技术与自动化装置] U691[自动化与计算机技术—控制科学与工程]