Reliability Analysis Of Electrical Control And Drive System Of Heavy-Duty CNC Machine Tools Under Epistemic Uncertainty

As an important processing equipment for large size parts,the heavy-duty CNC machine tool provides equipment support for the defense,military,aerospace and other areas about national security and people’s livelihood,and its manufacturing level is related to the development of manufacturing and national comprehensive strength.Under the advance of the national major projects,the technical level of CNC machine tools of our country has been improved obviously.However,compared with foreign mature machine tools,the domestic CNC machine tools exposed many problems in reliability such as the frequent failures in the operation process,the cause of the failure is difficult to find,the maintenance cost is high and the machining precision is low,therefore,it is urgent to start the reliability research for heavy-duty CNC machine tools.The role of electrical system is to control and drive the machine.As the key subsystem of the machine tool,whether the machine can run safety and reliably is related to the electrical system.Due to the lack of historical data,the limitation to carry out a large number of reliability experiments,the lack of information,the change of working environment,the complexity of system structure and fault mechanism,etc,there are a large number of uncertainties in the electrical system of CNC machine tools.The electrical system exhibits multiple performance states during operation.Common cause failure existed in the electrical system due to environmental and human factors.In view of the above problems,this paper devotes to the study of epistemic uncertainties,multi-state properties,common cause failures in the electrical system of the heavy-duty CNC machine tool,the details are as follows:(1)The heavy-duty CNC electrical system is divided into several subsystems based on the analysis of system working principle and function,which lays the foundation for the follow-up reliability analysis.Finding the common failure modes and the frequent failure subsystems through the fault analysis,which selected as the main research object of this paper.(2)The reliability analysis of the feed control system is based on fuzzy Bayesian Network which combined the fuzzy theory and the Bayesian Network.The traditional Bayesian Network is extended by the fuzzy theory,which solves the problem that the failure probability of traditional Bayesian Network must be exact value.In the view of the uncertainties in feed control system,the epistemic uncertainty of the root node failure probability is described by triangle fuzzy number of the fuzzy theory,the probability value in the conditional probability table of Bayesian Network is used to describe the uncertainty logical relationships between root nodes,which increased the ability of Bayesian Network to deal with uncertain information.(3)Using the method of combining evidence theory and Bayesian Network to analyze the reliability of the spindle drive system.Evidence theory quantify the epistemic uncertainty of the system by introducing the uncertain state focal element,expanding the state space of components by using the evidence theory,improving rules of traditional logic gates under the evidence theory,So Bayesian Network is extend by evidence theory.The results show that the method can not only deal with the system uncertain information well but also have high computation efficiency.The ? factor parameter model is used to deal with the problem of common cause failures in the system.(4)Bayesian Network can be used to analyze failure reasons.Finding the main cause of the system failure,which provides theoretical support for the failure detection and reliability growth of the system.