Research On Remaining Useful Life Prognostics And Health Management For Products With Degradation Processes

With the development of the aerospace industry, the high-speed railway transportation, the marine equipment and the heavy-duty machine tools and so on in China, the reliability of products with degradation processes influences the China’s national defense forces and all aspects of people’s livelihood. It demands research on the reliability theory of products with degradation processes. In total life cycle of products, there is an increasing request of health condition monitoring, real-time failure prediction and maintenance management, in order to maintain the health condition. Thus, this dissertation will concentrate on remaining useful life prediction and health management for products with degradation processes. Products show dynamical failure processes, long time cycle of preparation and transportation for maintenance resources, expensive maintenance costs, and difficult evaluation of failure crowdings and so on. The main idea of this dissertation is to study on the remaining useful life estimation with the degradation and shock models, to propose the remaining useful life estimation based on fatigue crack growth model and modified particle filter, to establish the optimal online maintenance decision makings framework with a scheduling threshold and a maintenance threshold according to prediction results, and to present an evaluation method based on near-extreme theory for failure crowdings.The main contents of this dissertation are summarized as follows:(1) To present a prognostic approach to estimate remaining useful life for systems subjected to degradation and shock processes. Degradation and shock processes are described as "elemental processes". Shock damage is the damage to a soft failure process caused by a shock process. Degradation damage is the damage to a hard failure process caused by a degradation process. Shock damage and degradation damage here represent the influence of elemental processes on compound processes. The soft failure process consists of a degradation process and shock damage, and the hard failure process consists of a shock process and degradation damage. Soft failure and hard failure processes are "compound processes", and either of them could lead to system failure. A remaining useful life prediction framework is established on the degradation and shock model, by using a particle filter to estimate the system condition and online distribution of remaining useful life with and without measurement noise in prognostics.(2) The fatigue crack growth in mechanical structures is an essential failure mode of products with degradation processes. By combing high-performance fatigue mechanics with filtering theories, the main study focuses on the fatigue crack growth of mechanical structures for diagnostics and prognostics. Fast & accurate deterministic analyses of fatigue crack propagations are carried out, by using the Finite Element Alternating Method (FEAM) for computing SIFs, and by using the Moving Least Squares (MLS) law for computing fatigue crack growth rates. Both the extended Kalman as well as particle filters are applied to estimate the crack lengths. Based on the estimated crack length, the estimation of remaining useful life based on fatigue crack growth is achieved, and a modified particle filter is proposed to drastically reduce the computational burden.(3) Considering the lead time for maintenance resources preparation, a prediction based optimal maintenance decision-making is established. A scheduling threshold is introduced to represent the starting of various maintenance resources preparation. The scheduling threshold, the maintenance threshold and the failure threshold construct an integral maintenance management scheme. By using this scheme, suppliers can schedule maintenance services in advance when the system condition reaches the scheduling threshold, and perform maintenance when the system condition exceeds the maintenance threshold. The long-run expected cost rate is the objective, which considers the maintenance cost based on system conditions, and considers the costs of the waiting time of suppliers and customers. Furthermore, the optimal maintenance plan is updated dynamically in the framework of Prognostics and Health Management.(4) When a group of identical products is operating by customers, if there is the aggregation of failures, it is more dangerous than just one failure. In order to produce reliable and safe products, it is critical to avoid the failure occurrences in a short time, and research on this phenomenon. The phenomenon of failures is measured by near-extreme system condition and near-extreme failure time. Near-extreme system condition and near-extreme failure time will be defined and calculated in offline solutions and online solutions, then an evaluation system of failures aggregation is established for both offline solutions and online solutions.