Philippe Weber

Overview on Bayesian networks applications for dependability, risk analysis and maintenance areas

In this paper, a bibliographical review over the last decade is presented on the application of Bayesian networks to dependability, risk analysis and maintenance. It is shown an increasing trend of the literature related to these domains. This trend is due to the benefits that Bayesian networks provide in contrast with other classical methods of dependability analysis such as Markov Chains, Fault Trees and Petri Nets. Some of these benefits are the capability to model complex systems, to make predictions as well as diagnostics, to compute exactly the occurrence probability of an event, to update the calculations according to evidences, to represent multi-modal variables and to help modeling user-friendly by a graphical and compact approach. This review is based on an extraction of 200 specific references in dependability, risk analysis and maintenance applications among a database with 7000 Bayesian network references. The most representatives are presented, then discussed and some perspectives of work are provided.

Complex system reliability modelling with Dynamic Object Oriented Bayesian Networks (DOOBN)

Nowadays, the complex manufacturing processes have to be dynamically modelled and controlled to optimise the diagnosis and the maintenance policies. This article presents a methodology that will help developing Dynamic Object Oriented Bayesian Networks (DOOBNs) to formalise such complex dynamic models. The goal is to have a general reliability evaluation of a manufacturing process, from its implementation to its operating phase. The added value of this formalisation methodology consists in using the a priori knowledge of both the systems functioning and malfunctioning. Networks are built on principles of adaptability and integrate uncertainties on the relationships between causes and effects. Thus, the purpose is to evaluate, in terms of reliability, the impact of several decisions on the maintenance of the system. This methodology has been tested, in an industrial context, to model the reliability of a water (immersion) heater system.

Reliability modelling with dynamic bayesian networks

Nowadays, the complex manufacturing processes have to be dynamically modelled and controlled to optimise the diagnosis and the maintenance strategies. The work reported here presents a methodology for developing Dynamic Bayesian Networks (DBN) to formalise such complex dynamic models. A small valve system is used to compare the reliability estimations obtained by the proposed DBN model and those obtained by the classical Markov Chain.

Evidential Networks for Reliability Analysis and Performance Evaluation of Systems With Imprecise Knowledge

This paper deals with evidential networks to manage imprecise probabilities. Evidential networks are directed acyclic graphs that handle random, and epistemic uncertainties thanks to Dempster-Shafer structures. After we recall useful bases of the Dempster-Shafer theory, and the relation with probability intervals, we explain how to handle imprecision. Evidential networks are extended with imprecise utility functions to deal with system performance evaluation problems. We explain the application of evidential networks to system reliability evaluation problems. Then, applications to multi-state performance evaluation are proposed. The last section of the paper is devoted to study case systems.

Bayesian Networks and Evidence Theory to Model Complex Systems Reliability

This paper deals with the use of Bayesian Networks to compute system reliability of complex systems under epistemic uncertainty. In the context of incompleteness of reliability data and inconsistencies between the reliability model and the system modeled, the evidence theory is more suitable to manage this epistemic uncertainty. We propose to adapt the Bayesian Network model of reliability in order to integrate the evidence theory and then to produce an Evidential Network. Three examples are proposed to observe the propagation mechanism of the uncertainty through the network and its influence on the system reliability.

Dynamic Bayesian Networks Modelling the Dependability of Systems with Degradations and Exogenous Constraints

Abstract The work reported here presents an original method to model dependability of systems, taking into account degradations and failure modes governed by exogenous constraints. The component degradation dynamics is considered as a serni-Markov process. Environmental behaviour introduces switching models conditioned by exogenous constraints. Dynamic Bayesian Networks (DBN) are employed to fonnalise such complex dynamic processes through a compact representation. DBN allow simulating these processes, taking into account events due to the environmental behaviour. A hydraulic system is used to illustrate the reliability estimations obtained by the proposed modelling method.

Reconfigurability analysis for reliable fault-tolerant control design

Reconfigurability analysis for reliable fault-tolerant control design In this paper the integration of reliability evaluation in reconfigurability analysis of a fault-tolerant control system is considered. The aim of this work is to contribute to reliable fault-tolerant control design. The admissibility of control reconfigurability is analyzed with respect to reliability requirements. This analysis shows the relationship between reliability and control reconfigurability defined generally through Gramian controllability. An admissible solution for reconfigurability is proposed according to reliability evaluation based on energy consumption under degraded functional conditions. The proposed study is illustrated with a flight control application.

Reconfigurable control design for over-actuated systems based on reliability indicators

Control allocation is a solution to distribute the control efforts among a redundant set. A new approach to manage the actuators redundancy in the presence of faults is proposed based on reliability indicators. The aim is to preserve the health of the actuators and the availability of the system both in the nominal behavior and in the presence of actuator faults. In degraded functional, a reconfigured control allocation strategy is proposed based on the on-line re-estimation of actuators reliability. A benefit of incorporate the reliability indicators on the over-actuated control system design is to manage smartly the redundant actuators and improve the safety of the system. The proposed approach is illustrated with a flight control application.

Design of a fault tolerant control system incorporating reliability analysis and dynamic behaviour constraints

In highly automated aerospace and industrial systems where maintenance and repair cannot be carried out immediately, it is crucial to design control systems capable of ensuring desired performance when taking into account the occurrence of faults/failures on a plant/process; such a control technique is referred to as fault tolerant control (FTC). The control system processing such fault tolerance capability is referred to as a fault tolerant control system (FTCS). The objective of FTC is to maintain system stability and current performance of the system close to the desired performance in the presence of system component and/or instrument faults; in certain circumstances a reduced performance may be acceptable. Various control design methods have been developed in the literature with the target to modify or accommodate baseline controllers which were originally designed for systems operating under fault-free conditions. The main objective of this article is to develop a novel FTCS design method, which incorporates both reliability and dynamic performance of the faulty system in the design of a FTCS. Once a fault has been detected and isolated, the reconfiguration strategy proposed in this article will find possible structures of the faulty system that best preserve pre-specified performances based on on-line calculated system reliability and associated costs. The new reconfigured controller gains will also be synthesised and finally the optimal structure that has the ‘best’ control performance with the highest reliability will be chosen for control reconfiguration. The effectiveness of this work is illustrated by a heating system benchmark used in a European project entitled intelligent Fault Tolerant Control in Integrated Systems (IFATIS EU-IST-2001-32122).

System Approach-Based Bayesian Network to AID Maintenance of Manufacturing Processes

The prospective work reported here explores a new methodology to develop Bayesian Network-based diagnosis and prognosis aids for manufacturing processes. This work is justified with the complex systems by the need of controlling and maintaining in dynamical way the global system performances in order to optimise the enterprise strategies. The added value of our methodology is to formalise the maintenance aid models from a priori knowledge both on the system functioning and malfunctioning by means of Bayesian Networks. The networks are built on adaptability principles and integrate uncertainties on the relationships between causes and effects. The feasibility of this methodology is tested in a manufacturing context with the maintenance aids on a lathe machine.