ABSTRACT
This paper presents the development and the application of a novel intelligent R&MM system by means of causal modeling. Such a R&MM system models the causal relationship of complex industrial systems in an generic way and assesses the reliability on a system level. Bayesian method is applied to model and integrate the data and information derived from individual parameters and components within complex systems. The intuitive modeling principle of Bayesian Belief Network (BBN), the mechanism of Bayesian inference and the intelligent reasoning ability enable the system to achieve the advanced R&MM. In the R&MM system, a generic Bayesian causal model and its reasoning framework are developed, which can deal with the R&MM for various complex industrial environments. Especially, the causal model takes the integration and reasoning on four R&MM levels, namely, the parameter level, the component level, the function level and the system level. The original concept of the integrated
1 INTRODUCTION
In industrial practice, a complex system is composed of a number of assembled components or subsystems. The reliability of the system relies on the combined reliability of the components. The malfunction of one of the components may lead to expensive process downtime if it causes a total shutdown of the system. Since an industrial system and relative production process cannot be designed to be 100% reliable and the reliability of system components decreases with aging and when degradation occurs, proper maintenance strategies and decision-making are required to maintain the reliability of the various components and the overall system. Therefore, reliability and maintenance management (R&MM) plays an important role to ensure the proper functioning and to achieve optimized performance of industrial systems.
