Research On Joint Optimization Of Production Lot And Maintenance Planning Assuming Operation-dependent Failure

Both production planning and maintenance arrangement are very important functions in the operation of a manufacturing corporation. Making decisions on production and maintenance separately, as is often done in practice and research literature, may not result in overall optimization. This paper aims to propose joint decision methods, respetively in a single-machine system and two multi-machine systems, that better integrate production planning and maintenance at the tactical level.Firstly, an integrated production lot and maintenance planning model for a singlemachine system is proposed. It considers operation-dependent failures and takes into account the loss of production time caused by both preventive and corrective maintenance. The aim is minimizing the sum of production costs and maintenance costs. For the model includes sophisticated nonlinear constraints, we propose an iteration-based solution algorithm. Computational experiments validate the effectiveness of the iterative algorithm and the superiority of our method compared with two other methods. The effects of several parameters on the results are also investigated.To be followed, the research has been extended to deal with the case of a multi-machine serial system, and imperfect preventive maintenance is taken into account together with the constraint of maintenance resources. The function relationship between the system production capacity and the maintenance decisions of single machines is investigated, starting from which the integrated model for the serial system has been eastablished. To deal with the large decision space of this problem, a heuristc solution-knowledge based algorithm and an adapted genetic algorithm are respectively proposed. The former one provides new perspective to the solution of an integrated production and maintenance planning problem, while the latter one is an elaborately adapted version of genetic algorithm for the integrated planning problem.At the end of the thesis, the problem of integrating production lot and maintenance planning in a multi-stage system with intermediate buffers is investigated. A hybrid algorithm based on decomposition of the problem and a metaheuristic solution procedure for the sub-problem is proposed, in order to deal with the high complexity of the problem.