中文摘要：

辊是使轧材产生塑性形变的主要工具，也是轧钢过程中的主要消耗部件，其合理的更换策略和库存设置对保证企业安全生产和降低生产成本至关重要。将同一生产线上多个轧辊组成的系统建模为一个同类可修多部件系统，结合轧辊本身的运行和维修特点，提出了成批更换及离线视情修磨相结合的轧辊更换修磨策略与备件库存订购策略的联合策略，在考虑确定的修磨时间，交付时间和非完美修磨效果的基础上。分析了生产过程中轧辊的成批更换、视情修磨和备件订购的相互制约关系，建立了以更换周期、报废阈值和库存参数为决策变量，以单位轧辊无限时间范围内的平均费用率最小为目标的仿真优化模型，最后采用离散事件仿真和遗传优化算法相结合的方法对模型进行求解。仿真结果表明，联合优化所得到的更换周期和库存参数更合理的，可以有效降低生产成本。

英文摘要：

Roller is an important tool for plastic deformation of rolling material and the main consuming parts in Rolling Industry. Reasonable replacement policy and spare parts inventory are essential to ensure production and reduce production costs. Taking the roller system in roll machine as a multi-unit with many identical units and combining the characteristic of roller, a joint optimization strategy of roll block replacement policy and condition- based repair and （S, s） type spare parts provisioning policy was proposed. Based on the assumptions of determinis tic maintenance time , lead - time and imperfect maintenance effect , the relationship between maintenance policy and spare parts inventory strategy was analyzed. On this basis, the simulation optimization model was estabhshed with trashed threshold and inventory parameter as the decision variable and minimum average cost rate in infinite time range as the target. The discrete event simulation and genetic algorithm were used to solve the model. The re- suits of simulation indicated that the joint optimization strategy could be got, and the proper retirement threshold and inventory parameter can decrease maintenance cost significantly.