Research On Dynamic Reliability Modeling And Assessment For Electromechanical System

As an important quality attribute of products, reliability will affect customer’s satisfaction and enterprise’s profit directly. Products are becoming more complicated in their structures and functions due to the continuous improvement of technology. Meanwhile, some particular electromechanical products work in dynamic working environment, which make their failure mechanism more complicated and also raise challenges to traditional reliability methods.Various dynamic reliability approaches have been proposed on traditional system reliability modeling methods, which are used to predict the failure characteristics varying with time as well as the influence of environment factors acting on products’ reliability. Selecting the electromechanical system in a particular engineering equipment as the object of study, the dynamic reliability modeling and assessment have been carried out in this thesis. The major work is as follows:(1) Reliability assessment based on improved Failure Mode, Effect and Criticality Analysis (FMECA)and Failure Propagation Model (FPM). The "Failure Cause" column in traditional FMECA sheet is extended to two columns, i.e. "Failure Cause" and "Cause Probability". Meanwhile, the Failure Propagation Graph (FPG) of the system failure modes is established based on FPM. The improved FMECA method is combined with FPG to conduct quantitative assessment of failure modes. The probability of each failure mode can be obtained by conducting fault diagnosis based on FPG and the cause failures’ probabilities as well as some basic components’ failure probabilities. Therefore, the Risk Priority Value (RPV) of each failure mode can be calculated and used as the reliability index to provide direction for maintenance policy making.(2) Multi-state reliability assessment based on Dynamic Bayesian Network (DBN). The Structure Analysis and Design Technology (SADT) and Failure Mode and Effect Analysis (FMEA) methods are used to analyze the function and failures of the system respectively. Taking the multi-state characteristic and status transition of system components into consideration, the DBN model of the system based on SADT and FMEA analysis results is established, which has the ability to conduct forward and backward reasoning at the same time. Through forward reasoning, the reliability of the ystem can be obtained; through forward reasoning, the reliability bottlenck can be recognized. A case study of pitching system is given out to verify the proposed method, and the result shows good consistency with practice.(3) Dynamic reliability assessment of electromechanical system considering environment influence based on Universal Generating Function (UGF). The reliability of a motor, which works under distance-varying vibration environment caused by road roughness, is studied. Taking the motor’s amplitude of vibration as the environment character, the function between the motor’s failure rate and vibration amplitude is established. On that basis, the system reliability is assessed by using UGF with dynamic parameters. In the model, the motor’s state value is not a certain constant, but an expression of vibration amplitude. Thus, the relationship between system reliability and vibration amplitude is established, and the influence of the vibration on system availability can be studied by analyzing the calculation results.The thesis has conducted some research work on reliability assessment of electromechanical systems by considering the influence factors including dynamic, multi-state and environment. The corresponding reliability models are established, and some case studies are also finished, which provide theoretic basis for dynamic reliability modeling and assessment of such kinds of systems.