Abstract
Abstract The purpose of this dissertation provides a dynamic preventive maintenance policy for the equipment inspected by sensors. The sensors can inspect several parameters of the equipment which can be generate a Health Index (H) in real time at each inspecting time nd, n=0,1,2,3…The H will be translated into equipment status and saved in the system database. According to the equipment status, we can decide the best preventive actions at time nd by the rule of the lowest total maintenance cost per cycle. The state of the equipment can be charge by H. Each state of the equipment has several preventive actions to choose. We suppose the state transition probability of the equipment as a steady state transition probability matrix, which means the state of the equipment is a Markov chain. Base on continuing updated H value, we can estimate the state transition probability of the equipment and the risk rate of each preventive action at every nd n=0,1,2,3…. As a result, the state transition probability of the equipment will be updated with nd. In brief, the major differences between this preventative policy and others are that this one policy is updated by time and the state transition probability of the equipment can be updated by time. Therefore, the enterprise can base on the equipment historical data to generate the updated transition probability with time, and follow the methodology as we mention above then it can build up an useful maintenance policy.Key word: Health index, Multiaction, Dynamic maintenance policy.