离散系统增量任务重调度多目标协同优化问题研究

Based on strategy the demand of energy saving and emission reduction, in order to satisfying the actual production requirement, it is studied that the rescheduling problem for positive or negative incremental tasks (positive, i.e., some new tasks have arrived, negative, i.e., some original tasks are cancelled) in discrete manufacturing system. It is solved that which tasks (may be original or new tasks) need to be scheduled and what schedluling will be implemented in rescheduling under adequate or limited resources. The objective of problem is to optimize the energy consumption and economic performance. In the heat treatment shop of quartz glass factory, the mathematical models merging mechanism are established for every problems adopting various ways. Besed on the advantages and disadvantages of the existing algorithms, double-deck algorithm, improved branch and bound algorithm and improved evolutionary algorithm are developed. The numerical experiments are designed to testing the performance of proposed algorithms. Following a case analysis for quartz glass factory is studied. Base on process and effect for solving above problem, the mathematical model and scheme of algorithm design are established for general rescheduling problem of incremental tasks. This research applies to solving practical problems of rescheduling for incremental tasks in dispersing manufacturing enterprise. The theoretical results of this research can enrich scheduling theory and lay a foundation for further studying of reshceduling for incremental tasks.

本项目从企业实际生产问题和节能减排的战略需求出发，研究了离散制造系统中，由于正负增量任务（正为新任务到达；负为初始任务取消）干扰需调整车间的初始调度方案，在满足资源约束的条件下，确定参与重调度计划的初始任务和新任务，制定考虑节约能源和经济效益的多目标协同优化问题的重调度方案。以某石英玻璃厂热处理车间为背景，融合机理分析运用多种方法建立数学模型；分析现有算法特点，基于问题特点和提出的问题性质设计双层求解算法、改进的分枝定界算法和改进的进化算法对问题进行求解；通过数值实验验证算法的有效性及性能，并进行案例分析。拓展分析问题的解决过程和效果，建立具有普适性的问题数学模型和算法求解方案，构建基于增量任务重调度的理论研究框架。本研究适用于解决离散企业中基于增量任务重调度的实际问题，研究成果可推广到其它具有弹性等待时间特点的重调度问题，丰富了重调度理论并为进一步研究增量任务重调度优化问题奠定理论基础。

本项目从企业实际生产问题和节能减排的战略需求出发，研究了离散制造系统中，由于增量任务干扰需调整车间的初始调度方案，在满足资源约束的条件下，确定参与重调度计划的初始任务和新任务，考虑在满足实际生产要求的前提下，以优化节约能源为目标制定调度方案的单机重调度问题。具体以某石英玻璃厂热处理车间为背景，分别建立了问题的一般优化模型、整数规划模型、动态规划模型和线性规划模型；根据问题和现有算法特点，基于问题特点和提出的问题性质，分别设计了动态规划算法、改进的分枝定界算法、Cplex求解整数规划模型和改进的遗传算法对问题进行求解；通过数值实验验证算法的有效性及性能，并进行案例分析。通过拓展分析问题的解决过程和效果，建立了针对此类问题具有普适性的数学模型和算法求解方案，丰富了基于增量任务重调度的理论研究框架。而且本研究有望应用到解决离散企业中基于增量任务重调度的实际问题，研究成果也可推广到其它具有弹性等待时间特点的重调度问题。