Diagnosis And Maintenance Management Optimization Study On Petrochemical Rotary Machinery

Rotary machinery is the core equipment in the petrochemical production. The failurefor any equipment will occur inevitably and maintenance is an essential means forensuring the normal operation of rotary machinery. But the fault reason of petrochemicalrotating machinery has the complexity, diversity and uncertainty characteristics, andmaintenance decision-making is influenced by many factors, so the process ofmaintenance decision-making management is very complex. For this reason, themaking-decision management on the petrochemical rotary machine was investigated inthis paper.The fault diagnosis model, which is based on Bayesian network, was established. Toprevent or avoid the recurrence of the similar fault again, first of all, the fault reasons areclassified and analyzed and the corresponding maintenance measures was put forward;Then the failure modes and effects analysis was combined with Bayesian networkmethods, so the weak link can be analyzed for the influence of the equipment operation. Apetrochemical enterprise of centrifugal pump was taken as an application example. Theresults show that the method can provide certain basis for fault diagnosis of thepetrochemical rotary machine fault diagnosis.The computer of equipment maintenance management system was developed. VBlanguage was used and modular design framework was designed in this system. It canmanagement the basic information of the equipment can be managed through it, and afault model database was built. Meanwhile the automation analyze for the Bayesiannetworks analyze was realized. The system can improve the efficiency of equipmentinformation management and the fault diagnosis.Taking the least equipment management cost within unit time as the target, amaintenance optimization model based on monotonous geometric process wasestablished，which can be applied to many types of life distribution. And Monte Carlosimulation method was introduced into this mode. The simulation program by Matlab wasobtained and can predict the optimum life cycle of the mechanical equipment of. The centrifugal pump for Weibull distribution of two parameters was analyzed as a study case.Results show that the model is reasonable, which can supply a scientific guideline formaintenance making-decision.