A Study On Statistical Quality Control-Condition Based Maintenance Method For Production Machine

The development and utilization of a large number of automatic and intelligent equipment is one of the important characteristics of the modern advanced manufacturing industry.The health and efficiency of equipment directly affect the quality and productivity,and thus determine the manufacturing cost and profit of the factory.After digital transformation,manufacturing enterprises in our country have acquired the ability to collect and store the locale information.According to the specific production mode and historical operation data of the factory,the customized equipment maintenance strategies and schemes is helpful to further improve the management level and comprehensive competitiveness of the factory.This paper proposes a method called Statistical Quality Control-Condition Based Maintenance(SQC-CBM).In this method,the fluctuation of equipment processing quality is taken as the evaluation index of equipment health status.According to the theory of statistical quality control,the operation status of equipment is categorized as “under control” and “out of control”,which can be indentified by the control chart.The maintenance is then performed when the control chart alarms.With reasonable simplification and assumption for the batch production process,this paper discusses the possible states of the equipment under the specific production and maintenance strategy,and analyzes the state transition probabilities or rates according to the historical data as well as the control chart and Bayesian theory.Then the SQC-CBM Markov state-space model is built.This model is capable to calculate the expected input or output of the production system under the corresponding process quality control and equipment maintenance strategy,so as to obtain the optimal solution for the lowest expected cost or the highest expected profit.The SQC-CBM method for sampling inspection by attributes is modeled by Markov chain with discrete time and state.Under the assumption that the time of sampling inspection and equipment maintenance are negligible,the processing,inspection and maintenance costs of the serial production system under the monitoring of the attribute control charts are analyzed and calculated.Then,parameters for the optimal attribute control chart can be obtained so as to guide the condition based maintenance.For the measurement sampling test,this paper first establishes the continuous time homogeneous Markov model of a single station,which is verified by the Tecnomatix simulation.In order to reduce the shutdown loss caused by false alarms and to reserve necessary preparation time for the condition based maintenance,this paper proposes a delayed maintenance strategy,which allows the equipment to continue to operate for a period of time,instead of shutdown immediately,when the control chart alarms.The established model is used to solve and analyze the optimal equipment maintenance policy under different production conditions.Consequently,the SQC-CBM recursive model of serial production system is further established.This model analyzes the transfer effect of the outage and defective outflow on the downstream station under the assumption of infinite inventory capacity.The optimal SQC-CBM decision parameters are searched by the genetic algorithm,to systematically optimize the maintenance strategy for each machine.Finally,this paper introduces the SQC-CBM Decision Support System(DSS),which is developed to optimize the maintenance strategy for the production equipments,and to provide effective SQC-CBM solutions by making good use of the massive historical data in the information system.Several engineering application cases are provided to verify the feasibility of the system.