Research On The Reliability Evaluation And The Maintenance Optimization Method Of The Equipment

In recent years, with the fast development of science and technology, engineering facilities are growing in the direction of integration, intelligence, accuracy and automation. Plant maintenance has become one of the key factors in guaranteeing business’success. The study on plant maintenance and fault diagnosis centering on reliability has great importance in ensuring facilities’safe and continuous operation.The present paper is funded by the projects of Study on Pattern Theory of Plant e-maintenance and Supporting Technology (Project number 70601010) and Sparse Signal Decomposition Based on Multi-scale Chirplet and Its Application to Mechanical Fault Diagnosis (Project number 50875078) in National Natural Science Foundation of China. The main researches and achievements are as as follows:1. Aiming at the reliability evaluation and optimization of the continuously operating equipment in process manufacturing enterprise, a reliability evaluation and optimization method based on discrete event modeling and simulation is proposed in this paper. Taking the long-term operation of the equipment as a discrete event system, the present paper sets up a process model of the equipment running on the platform of Plant Simulation, whose key parameters are defined by parameter estimation based on device’s historical operating record. The long-term simulation running of the model is used to evaluate the operating reliability of the equipment, and the reliability can be upgraded through the optimal allocation and improvement of maintenance resources, which provides technical support for the plant’s long cycle running with high load. Validity and feasibility of the method are verified by the successful application in some petrochemical company.2. In view of the current reliability-centered maintenance methods not covered in the analysis because of lack of reliability, as well as the maintenance of human reliability analysis due to the importance of a consideration because of the reliability of reliability-centered maintenance improve the method, because of the reliability of analytical techniques into the reliability-centered maintenance analysis, in order to safeguard the system, the development of enterprises to provide a more comprehensive and effective implementation of the theory and methodology support. This method is a successful application of petrochemical enterprises proved its effectiveness. 3. A new method for adaptive zero phase time-varying filter design and its application in speed-changing gearbox fault diagnosis is proposed in the present paper. Aiming at the problems in time-varying filter design based on Multi-scale chirplet sparse signal decomposition, where there is phaseshift which will cause signal distortion, an adaptive zero phase time-varying filter design is proposed. The filter designed filter single component AM-FM signals from non-stationary multi-component signals, onto which Hilbert envelope analysis is carried out. Hilbert envelope analysis’single component requirement is satisfied. The method is applied to speed-changing gearbox fault diagnosis.4. Aiming at the weaknesses of Particle Swarm Optimization algorithm (PSO) such as the easiness to get trapped into local optimal, slow convergence speed, and low precision, an improved PSO algorithm called Selectivity-Particle Swarm Optimization algorithm (S-PSO) is proposed in this paper. Differently from traditional PSO, according to S-PSO particles can refer not only to the optimal strategy, but can also select a better strategy based on their own consideration and then imitate it. When selecting local better strategy, S-PSO is introduced to increase particles’intelligent behaviors. Based on the authority of strategy, the overlay group of strategy and spatial distance between particles themselves and the overlay group, particles will select the local better strategy. Experimental simulation results show that the global searching ability and convergence speed and precision of S-PSO algorithm is much better than traditional PSO.Apply S-PSO onto the solving of optimal model of equipment preventative maintenance cycle. The results show that S-PSO enjoys strong overall searching ability. With relatively fast speed, it will converge at the global optimum, which will provide strategy and information for companies’s actually maintenance.