研究了单晶圆厂、多针测厂、多封装厂、多终测厂的半导体供应链协同计划问题,建立了该问题的数学模型,提出了用于求解上述问题的知识进化算法方案。对典型算例进行了仿真并与粒子群算法进行了比较,结果表明了半导体供应链协同计划模型及算法的有效性,同时也表明了知识进化算法取得了比粒子群算法更好的优化效果。
The problem of coordinated planning for semiconductor manufacturing supply chain, comprised of single wafer fabrication, multiple wafer probing factories, multiple assembly factories, multiple final test factories, is studied. The model of this problem is constructed and solved using a method based on Knowledge Evolution Algorithm. The detailed implementation of the method is illustrated, with an example. By comparing the results, it is shown that coordinated planning remarkably reduces the supply chain cost and Knowledge Evolution Algorithm has higher searching efficiency than the particle swarm optimization algorithms.