Research Of Predictive Mainteannce Policy And Integrated Production Scheduling Model Based On Machine Degradation

With the development of industrial technologies in manufacturing enterprise, it is vital to ensure the machine reliability for smooth production. Maintenance management, as an important part in manufacturing system, is the basic and key to ensure equipment productivity and reduce manufacturing cost. Although there are some studies on maintenance policies, most of them are time-based maintenance policies, which ignores the influence of machine degradation. It is thus essential to study machine operation and degradation to develop predictive maintenance policy. Moreover, by considering the influence of maintenance operation on production scheduling in manufacturing system, maintenance operation should be integrated into the traditional production scheduling models. Due to machine degradation, a production scheduling model integrated with predicitve maintenance is proposed. This paper studies a single repairable machine and its production process. Based on the available research on maintenance policies and production scheduling, this paper discusses machine’s operation process and describes machine degradation, which greatly supports predictive maintenance planning. In addition, this paper integrates predictive maintenance into production scheduling subject to machine degradation, which helps extend traditional production scheduling research. In this paper, a performance assessment method of the machine is studied first. Machine’s health index (HI) is estimated based on statistical pattern recognition (SPR). Then, with machine’s HI, machine’s remaining maintenance life (RML) prediction model is built, which can well describe machine’s deterioration process. Later, according to machine’s RML, one predictive maintenance policy is developed. At last, by considering the influence of machine degradation on production scheduling, a production scheduling model integrated with predictive maintenance is proposed, which can optimize real production process.Based on the previous research and literature, this paper researches on predictive maintenance policy and integrated production scheduling model based on machine degradation, which focuses on the following parts:(1) Machine’s HI is estimated based on SPR method. Then, machine’s RML prediction model is built with machine’s HI.A complete machine’s performance assessment procedure is established based on SPR. In this procedure, the collected data is analyzed by data pre-processing, feature extraction and feature dimension reduction methods. Based on SPR concept, machine condition is identified. Chi-square test is used to estimate machine’s HI to help describe machine’s real health condition. With the machine’s HI information, machine’s RML is defined and machine’s RML prediction model is constructed, which helps describe machine degradation.This RML prediction model extends the applications of machine’s HI and supports predictive maintenance policy.(2) By considering machine degradation, predictive maintenance policy based on machine’s RML is developed.Considering machine degradation, a predictive maintenance policy is developed based on machine’s RML. This predicitve maintenance policy studies the influence of machine degradation to avoid the situations of low machine reliability and high manufacturing cost caused by previous failure-based or time-based maintenance policies. The maintenance level of machine’s HI and predictive maintenance cycles are the two decision variables.In this research, machine’s RML of each maintenance cycle is predicted and optimized. By considering the interaction of maintenance operation and machine performance, the maintenance cost and the operational cost should be variable if machine’s condition is changed. With the development, this predictive maintenance policy is enhanced to satisfy real manufacturing process.(3) An integrated production scheduling model considering predictive maintenance is proposed based on machine degradation.As most previous production scheduling research ignores machine reliability and assumes machine is available all the time, they cannot satisfy real manufacturing process due to machine degradation. This paper studies the influence of machine degradation on production scheduling and the influence of maintenance operation on production scheduling. Based on machine’s RML prediction model, this paper proposes a production scheduling model integrated with predcitve maintenance, which helps extend the traditional production scheduling research.With machine’s RML prediction model, machine’s RML is predicted to describe job’s completion time and tardiness. Then, with the aim to minimize the maximum tardiness of jobs, the optimal job sequence with predivtive maintenance planning is obtained. This integrated production scheduling model can ensure both machine reliability and productivity, which could be taken as new theory and practical references for production scheduling research.This research work can give qualitative and quantitative description of machine degradation, and provides a predictive maintenance policy considering machine degradation, finally achieves an improved production scheduling model integrated with predictive maintenance. This research can help enterprise ensure machine reliability, reduce manufacturing cost, increase productivity and enhance its competitiveness. It can provide great support to develop maintenance management and achieve the full performance of manufacturing systems, and provide scientific guidance for digital production management.