Systems health monitoring is essential to guaranteeing the safe, efficient, and reliable operation of engineering systems. Integrated systems health management methodologies include fault diagnosis mechanism. Diagnosis involves detecting when a fault has occurred, isolating the true fault, and identifying the true damage to the system. This important issue is even harder when the systems to be diagnosed are dynamic and spatially distributed systems with their successively increasing complexity. For such systems, a single diagnostic entity having a model of the whole system approach is inappropriate. Whereas a distributed approach of multiple diagnostic agents can offer a solution. An overall systematic solution for these issues could be obtained by an artificial intelligent mechanism called the multi-agent system (MAS). This paper presents a MAS model for fault diagnosis based on logical theory of diagnosis. In this approach, each local diagnostic agent has knowledge above its subsystem and an abstract view of the neighboring subsystems and it is able to determine the local minimal diagnoses that are consistent with global diagnoses. The multi-agent models are simulated in Java Agent Development Framework and are applied to the preheated cement cyclone in the workshop of SCIMAT clinker.

No abstract

Fault prognosis in industrial plants is a complex problem, and time is an important factor for the resolution of this problem. The main indicator for the task of fault prognosis is the estimate of remaining useful life (RUL), which essentially depends on the predicted time to failure. This paper introduces a temporal neuro-fuzzy system (TNFS) for performing the fault prognosis task and exactly estimating the RUL of preheater cyclones in a cement plant. The main component of the TNFS is a set of temporal fuzzy rules that have been chosen for their ability to explain the behavior of the entire system, the components’ degradation, and the RUL estimation. The benefit of introducing time in the structure of fuzzy rules is that a local memory of the TNFS is created to capture the dynamics of the prognostic task. More precisely, the paper emphasizes improving the performance of TNFSs for prediction. The RUL estimation process is broken down into four generic processes: building a predictive model, selecting the most critical parameters, training the TNFS, and predicting RUL through the generated temporal fuzzy rules. Finally, the performance of the proposed TNFS is evaluated using a real preheater cement cyclone dataset. The results show that our TNFS produces better results than classical neuro-fuzzy systems and neural networks.

No abstract

No abstract