考虑技术突破与技术升级的能力规划问题研究

技术进步与市场需求不确定性,都是制造业能力规划问题需要面对的最重要因素.以往研究仅考虑技术突破而没有考虑技术升级,较多考虑能力扩张而较少考虑能力更替,本文首次提出了一种同时考虑技术突破和升级的设备采购与替换的集成能力规划模型.在需求和技术进步双源不确定性条件下,采用Scenario方法建立了多计划期能力规划的集成决策模型,探求能力扩张、替换、维护的期望总成本最低.面对该决策问题所建立的非线性混合整数规划模型,本文为该NP complete问题设计了一种基于遗传算法框架的启发式求解算法,即通过增加变换操作的方法将该问题转换为可以随机进化求解扩张方案最优化求解替换方案的一个等价问题.在种群初始化过程中采取仅选择能力扩张决策进行染色体部分编码策略,然后用经典优化方法针对每个个体精确求解最优能力替换决策,将扩张和替换整体成本作为个体适应度参与个体评价与种群进化.实验结果表明,技术升级在不同需求变化情景下都可以有效降低能力规划的成本,且本文提出的启发式算法对于求解此类规划问题具有很好的收敛稳定性和性能稳定性.

Capacity planning with technological breakthroughs and upgrades

Technological progress and uncertainty of demand are important factors of problems of capacity planning in manufacturing. The majority of previous studies did only focus on technological breakthroughs rather than technical upgrades, two different ways of the technological progress, and paid more attention on capacity expansion while less attention on replacement. This paper presents for the first time an integrated capacity planning model that takes into account both technological breakthroughs and upgrades for equipment procurement and replacement. Under stochastic demand and technological progress, we use scenario-tree method describe uncertainty of two factors above, and propose an integrated model of multi-period capacity planning problem with the objective of minimum expected total cost of capacity expansion, replacement and maintenance. Faced with the nonlinear mixed integer programming model established by the problem above, the paper designs a heuristic solution method based on genetic algorithm (GA) framework for this NP complete problem, which converts the problem into an equivalent problem by performing certain translations. The equivalent problem can be solved by GA and classic optimization. In the process of population initialization, only expansion decision variables are chosen as the evolutionary population in GA, and then we use the classical optimization method to accurately solve the optimal capacity replacement problem for each individual, and use the total cost of expansion and replacement as the individual fitness for individual evaluation and population evolution. A numerical experiment illustrates that the technological upgrades can effectively reduce the cost of capacity planning under different demand scenarios, and the heuristic algorithm proposed in this paper has good convergence stability and performance stability for solving such planning problems.