Research On Integrated Optimization Model Of Statistical Process Control And Maintenance Decision

Statistical process control and maintenance decision are two typical research fields which aiming at ensuring the product quality and improving system reliability.The former one focuses on making optimal monitoring plan in order to detect the quality shift as soon as possible and prevent high cost from defective products.While the latter one aims at making reasonable maintenance schedules in order to prevent production interruptions from the system breakdown.These two aspects have close relationship with each other and have a significant influence on the production cost.By analyzing the existing literatures comprehensively,this dissertation focuses on the production process and establishes the economical model by integrating SPC with maintenance decision.Based on the renewal process,the stochastic process and the system reliability theory,the inner relationship between different essential production factors are studied.Including but not limited to the function type of the quality shift,the failure mechanism,the selection of control chart and the calculation of average run length,the related issues of the integrated model is researched.The main contents of this dissertation is as follows:Considering the quality shift is a function of the time,an integrated model with SPC and maintenance decision is proposed based on the simple failure mechanism of mean value shift.Then the mean control chart is conducted to monitor the quality characteristic.If a true alarm is generated,the predictive maintenance is utilized to restore the system as new.If no true alarm is generated in a certain period,the planned maintenance is utilized to restore the system as new.Based on the above assumptions,the basic structure of the integrated model is established.This dissertation analyzes the increasing type of quality shift influence on the statistical performance and the out-of-control run cost.Then the mathematical model is established aiming at minimizing the average run cost and solved by the tabu search algorithm.Moreover,the economical performance of the integrated model is compare with the pure SPC model and the pure planned maintenance model when the quality shift is the linear function,the quadratic function and the exponential function.Considering the production system consists of product process and equipment,a delay-monitor-based model is proposed by integrating the SPC and the maintenance decision.After the interact relationship between production process and equipment is analyzed,a hybrid failure mechanism including the quality shift and equipment failure is proposed.The mean control chart based on the delayed monitoring strategy is proposed.Combing with four different maintenance methods,an integrated model including ten production scenarios is established based on the fixed shift.The Monte Carlo simulation method is conducted to validate the practicability of the proposed model.The model is solved by the genetic algorithm and then compared with another integrated model with whole process monitoring.Considering the stochastic distributed quality shift,the optimization of the economical-statistical EWMA control chart is studied.A renewal process consists of in-control process,out-of-control process and maintenance process is proposed.A quality loss calculation method based on the probability that the quality shift exceeds the specification limits is proposed,and the EWMA control model is build aiming at minimizing the average run cost.A Markov method is proposed to calculate the average run length of the uni-EWMA control chart which includes multiple observations.Then,the economical-statistical uni-EWMA control chart is designed with the statistical constraint of the average time to signal.The uni-EWMA model is designed and solved by the genetic algorithm.What's more,the typical mean control chart and the classical EWMA control chart is compared with the proposed model under different shift distribution functions.Considering the hybrid failure mechanism which includes quality shift and equipment failure,the multi-EWMA control chart model is established for the multiple quality characteristics and small quality shift.A Markov method is proposed to calculate the average run length of the multi-EWMA control chart which includes multiple observations.Then,the economical-statistical multi-EWMA control chart is designed and solved by the particle swarm optimization.The result is compared with that from the pure economical multi-EWMA control chart.Based on the above research findings,a prototype system integrating SPC and maintenance model is developed.The monitoring and maintenance strategies are validated by conducting a real case in an engineer manufacturing factory.Finally,the summation of this dissertation is made,and the author points out several further directions on this issue.