Moncef Tagina

Reliability data for improvement of decision-making in Analytical Redundancy Relations Bond Graph based diagnosis

The method of Bond Graph based Analytical Redundancy Relations in Fault Detection and Isolation is explicitly associated with components faults, this is due to architectural and functional aspect of the Bond Graph tool. This allows using the reliability of each component to improve the decision-making step. The purpose of this paper is the improvement of the classical binary method of decision-making, so that it can treat unknown and identical signatures of failures. This approach consists of associating the measured residuals and the components reliability data to build a Hybrid Bayesian Network. This network is used to determine the posterior probabilities of the failures. As application, the approach is simulated on a controlled two-tank system.

Causal approaches and fuzzy logic in FDI of Bond Graph uncertain parameters systems

In this paper, two isolation methods issued from FDI and AI approaches are compared. These methods are interested with bond graph (BG) modeled uncertain parameters systems. Residuals are generated from the Diagnostic Bond Graph (DBG). Detection is based on fuzzy logic approach and its different parameters are determined off-line. By exploiting the causal properties of the BG model, two isolation methods are proposed: Fault Signature Matrix (FSM) and causal paths covering of the causal graph. A simulation example is provided to show the performances of the proposed methods.

Diagnosis of critical embedded systems: application to the control card of a railway vehicle braking systems

Embedded systems are more and more used to control critical systems. In this paper, we propose a diagnostic approach to increase the security of control of critical embedded system based on digital components. This is a part of a study to design of an electronic card to control a railway vehicle braking system. Because of the critical aspect, it is necessary to diagnose the failures of the control card to process them online for safety purposes. In this paper, we propose to use diagnoser technique based on timed automata. But since this technique suffers of combinatorial explosion and because digital devices are characterized by a lot of input/output, our approach proposes to make an abstraction of the system behavior to reduce the size of the models and to implement a kind of distributed diagnosers.

Consistency's checking of chronicles' set using Time Petri Nets

In DES, there are two basic approaches to diagnosis: the first approach is the diagnosers [1] and the second approach is the chronicle [2]. The first approach has limitations including the issue of combinatorial explosion. On the other side, it offers tools to study the diagnosability of the models constructed. The chronicles are easier to write but pose the problem of the guarantee of the completeness and especially the consistency of a given base. The objective of this work is to propose a method to verify the consistency of a set of chronicles from a rewritten form of T-Time Petri Nets (TPN).

Robust fault detection and isolation in bond graph modelled processes with Bayesian networks

The main objective of this paper is to present a new method for Fault Detection and Isolation (FDI) of non-linear uncertain parameters systems modelled by bond graphs (BGs) with Bayesian networks (BN). From the BG model of a process, residuals, which are fault detectors, are determined directly from the Diagnostic Bond Graph (DBG). In ideal conditions, those residuals are equal to zero. But in practice, owing to uncertainties, perturbations and measurement noises, residuals are different from zero. Classical approaches used thresholds to deduce whether a process is in normal operating mode or in faulty mode. In our approach, we generate a statistical decision procedure to detect the operating mode. For isolation, a Bayesian network is generated by covering the causal paths of the DBG, and the method proposed by Weber et al. is exploited. A simulation example on a three tanks system is provided to show the efficiency of the proposed FDI procedure.

From a functional graph to a T-timed Petri Nets model for the diagnosis of complex system

In this paper, we are interested in the monitoring of complex systems modeled by functional graphs. We propose a diagnostic algorithm to isolate and identify the causes of failure located at a node of the functional graph. We establish hypotheses, first, on the choice of type of function who causes failure, and second, on the choice of the number of function who are directly observable. The key contribution of our approach is the establishment of behavioral models based on T-timed Petri Nets to refine the diagnosis based on functional graphs and to finally isolate faults. This study proposes first a generic model for representing the behavior of nodes in the functional graph according to a number of rules. These rules are designed to combine the partial models to form a global model based on a functional graph structure.

Application of artificial intelligence in Fault Detection and Isolation of uncertain parameter systems

In this article, two original intelligent methods for Fault Detection and Isolation (FDI) affecting sensors and actuators in case of bond graph modelled uncertain parameter systems are proposed. First, a fuzzy approach based on residual processing is proposed for FDI offline, binary approach and fuzzy approach are compared through an illustrative example. Secondly, a multilayer perceptron trained with resilient backpropagation algorithm is proposed for FDI online.

Automatic Acquisition and Update of a Causal Temporal Signatures Base- for Faults Diagnosis in Automated Production Systems

Causal Temporal Signatures (CTS) is an efficient formalism for behaviors description and recognition of fault diagnosis in Discrete Event Systems (DES). The main advantages of this formalism are the readability and the expressivity. Indeed, it is able to describe clearly all desired behaviors and it is understandable and readable by an expert in the field. However, it raises the problem of acquisition and updating of expert knowledge stored in a CTS base. In this paper, we suggest an incremental learning approach based on the simulation to acquire and update automatically a consistent CTS base. The proposed approach is illustrated with an example applied to the turntable helps to understand the different modules of the method.

A Multi-agent Architecture for Sensors and Actuators’ Fault Detection and Isolation in Case of Uncertain Parameter Systems

In this work, a multi-agent system for fault detection and isolation is proposed in case of faults affecting sensors and actuators of bond graph modelled uncertain parameter systems. The diagnostic system is composed of different agents cooperating and communicating through messages exchange, each agent is specialized in a specific task of the diagnosis process. Detection and isolation routines are based on residuals processing. Residuals are obtained from bond graph model. To illustrate our approach, a hydraulic system in case of 5% of uncertainties is simulated under Matlab/Simulink environment. The diagnostic system is implemented in JADE environment and has shown good detection and isolation results.