Maintenance optimisation in a cement industry raw-mill system using genetic algorithm

This paper describes a maintenance optimisation framework for deriving optimal maintenance schedule for a process plant using genetic algorithm (GA). The two possible alternatives considered are: imperfect maintenance and replacement. The maintenance model incorporates both the corrective maintenance and preventive maintenance actions. GA search heuristic is used to optimise the choice of maintenance or replacement to achieve the minimum cost with target reliability. The model is evaluated by Monte Carlo simulation in terms of present value of the cost. The flexibility of the Monte Carlo method allows the inclusion of several practical aspects such as deteriorating repairs, aging and service variations. This study is carried out in the raw-mill section of a cement industry in Tamil Nadu, India.