Joint Optimization of On-Line Decision-Making for Maintenance and Buffer Allocation for Serial Production System

Abstract:

Aimed at the problem that the degradation state of equipment cannot be obtained online for serial production system， a joint optimization model of maintenance、quality control and buffer allocation was proposed. Based on the quality index of the hidden Markov degradation system， an online decision-making strategy for maintenance was proposed. The recursive formula of processing time and completion time was derived， and the buffer allocation model with limited buffer capacity was established. In order to minimize the total cost， a joint optimization mathematical model to decide the optimum quality threshold for maintenance and buffer allocation was established. The Monte Carlo simulation algorithm based on machine state-transition was used to estimate the expected cost of the system and a tabu search algorithm with cellular automatic mechanism neighborhood rules was proposed to solve the model. The numerical experiments verify the validity of the proposed joint optimization model and algorithms.