Selecting optimum maintenance activity plans by a unique simulation-multivariate approach

摘要

This study presents an integrated computer simulation-multivariate approach for selection of optimum maintenance activity plans. First, maintenance activities are modelled by computer simulation. Production and maintenance functions are estimated using historical data. Then, simulation is carried out for different scenarios, which are combinations of periodic maintenance and number of maintenance crew. Several outputs including machines and operators' availability, reliability, efficiency and queue length are computed. Four multivariate methods, namely data envelopment analysis (DEA), artificial neural network (ANN), principal component analysis (PCA) and numerical taxonomy (NT) are used to select the optimum policy. Finally, statistical methods are used to select the most reliable method for selecting the optimum scenarios.