Nowadays systems usually consist of multiple machines, whether type or scale. Some of the machines face an increasing rate of erosion through their life increasing. However, some issues are still needed to be discussed. On one hand, a lot of preventive maintenance models have been developed for single machine. On the other hand, in some of preventive maintenance models, the PM intervals are considered constant for all of the periods. Also most of developed maintenance models are focused on the cost.The preventive maintenance activities resulted in a certain cost and time that often required to stop the production line. In this case, it would be better to provide a model for coordination of production planning and maintenance.In this research, development of a model which is suitable for determining the dynamic (considering machine status in each cycle) optimal interval for preventive maintenance operations is considered in series - parallel systems by using of multi objectivedecisions methods to make a cost-effective system for whole-system maintenance. Also for reduction of stopping the production line, assuming the buffer is added to the model and the optimal buffer size will be determined by the model. In this case, No damages will be made in production planning while performing maintenance operations on each machine. Matlab 1 software is used for optimization. At the end, performance of the method is demonstrated by solving a sample problem and after that sensitivity of the model versus parameters and results are measured.
