不确定环境下复杂产品维护、维修和大修服务资源调度优化

基于对复杂产品维护、维修和大修（MRO）协同服务资源调度的需求分析，从现实角度出发，建立资源调度时间和服务执行时间参数不确定条件下的随机机会约束规划数学模型. 提出由随机模拟、神经网络和离散粒子群优化算法组合成的混合智能算法，求解所提出的优化问题. 随机模拟方法为所建立的神经网络模型提供训练样本集，得到的训练样本集被用于训练神经网络模型以逼近优化目标函数，训练后的神经网络模型被用于代替优化目标函数来执行粒子群算法优化迭代. 该混合算法能有效提升时间参数不确定条件下的复杂产品MRO协同服务资源调度双目标优化问题的求解速度. 案例分析表明，相比于确定性条件下的优化算法，所提出的随机机会约束规划模型和混合算法更适用于求解现实中不确定条件下的MRO服务资源调度问题，所求得的调度方案在实际执行中具有更好的鲁棒性.

Maintenance,repair and overhaul/operations service resource scheduling optimization for complex products in uncertain environment

A stochastic chance-constrained programming mathematical model under uncertain resource scheduling time and uncertain service execution time was built from a realistic perspective, based on the requirement analysis of the collaborative maintenance, repair and overhaul/operations (MRO) service resource scheduling of complex products. A hybrid intelligent algorithm composed of the stochastic simulation, the neural network and the discrete particle swarm optimization algorithm was proposed to solve the proposed optimization problem. The training sample set produced by stochastic simulation was used to train the neural network for the approximation of the optimization objective function. The trained neural network model was used to replace the optimization objective function to perform the optimization iterations of particle swarm algorithm. This hybrid intelligent algorithm can effectively improve the solving rate of bi-objective problem of collaborative MRO service resource scheduling of complex products under uncertain time variables. The results of case study showed that the proposed stochastic chance-constrained programming model and the hybrid intelligent algorithm were more suitable for solving the MRO service resource scheduling problem under uncertainty in the reality, compared with the optimization algorithm under certainty. The proposed scheduling scheme was more robust in practical implementation.