Belkacem Ould Bouamama

Bond Graph Model Based For Robust Fault Diagnosis

This paper deals with residuals and thresholds generation in presence of parameter uncertainties, using only one bond graph representation in linear fractional transformation form (LFT), applied for mechatronic system. The residuals sensitivity analysis based on the fault detectability indexes, with Interval and Statistical methods are used for their evaluation. The presented methodology is applied on electromechanical test bench in order to detect a backlash phenomenon. Experimental tests show the influence of the dead zone magnitude and system parameters variations on the residuals.

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.

Pseudo bond graph for fault detection and isolation of an industrial chemical reactor part II: FDI system design

This second part of the two series papers deals with the problem of bond graph model based fault diagnosis for a class of chemical reactions taking place in a jacketed reactor. A bond graph model based fault detection and isolation approach applied to design of a supervision system for a complex chemical reactor is presented. The nonlinearities and energetic couplings involved in the dynamics require a suitable supervision environment. Furthermore, the causal and structural properties of the bond graph tool are used to design Fault Detection and Isolation (FDI) algorithms (i.e. the generation of fault indicators) to determine the availability of the chemical process. Thus, a unified framework is developed to take into account faults affecting sensors, actuators and process. Structural analysis of the model is utilized to obtain the fault signatures for fault isolation. The decision procedure to detect and isolate the faults uses the residuals, evaluated online. A detailed kinetic modelling approach had been proposed in previous work to simulate chemical reactions and analyze the appearance of secondary reactions under industrial conditions in this present paper.

Sensor fault detection of energetic system using modified parity space approach

The main purposes of this paper concern residual generation based on analytical redundancy relations for sensor fault detection and isolation of steam generator. These relations are used to detect and isolate sensor fault based on the elimination of the unknown variables of the system. The nonlinear system of steam generator modeled in state space form is linearized around operating point linear analytical redundancy relations are generated by using parity space technique through individual observability of each sensor. This approach will be called modified parity space approach. The residuals computed on line show the effectiveness of this method.

Sensor location with respect to fault tolerance properties

This paper deals with the sensor location of non-linear systems with respect to fault tolerance properties. Non-linear observability and individual observability indices are introduced and used to build minimal and redundant sensor sets. These sets are organised into an oriented graph which contains all the possible reconfiguration paths for which the estimated variables remain observable. The evaluation of fault tolerance is based on fault tolerance properties: a structural and a probabilistic criteria, for which observability properties remain satisfied. The developed methodology is applied to a steam generator installation which represents complex non-linear system widely used in process engineering.

Design of Supervision Systems: Theory and Practice

The term “supervision” means a set of tools and methods used to operate an industrial process in normal situation as well as in the presence of failures or undesired disturbances. The activities concerned with the supervision are the Fault Detection and Isolation (FDI) in the diagnosis level, and the Fault Tolerant Control (FTC) through necessary reconfiguration, whenever possible, in the fault accommodation level. The final goal of a supervision platform is to provide the operator a set of tools that helps to safely run the process and to take appropriate decision in the presence of faults. Different approaches to the design of such decision making tools have been developed in the past twenty years, depending on the kind of knowledge (structural, statistical, fuzzy, expert rules, functional, behavioural…) used to describe the plant operation.How to elaborate models for FDI design, how to develop the FDI algorithm, how to avoid false alarms, how to improve the diagnosability of the faults for alarm manage...

Monitorability Indexes and Bond Graphs for Fault Tolerance Analysis

: nThe safety of complex industrial processes such as chemical or power processes is of prime interest. In this paper, the monitorability indexes are introduced to build a monitoring instrumentation architecture assuring a predefined level of safety. Easily calculated from the bond graph model, these indexes give a measurement on how much the instrumentation system can be efficient in spite of a decrease in the performance of the process information system related to some faulty sensors. An example illustrates our approach.

MONITORABILITY INDEXES AND BOND GRAPHS FOR FAULT TOLERANCE ANALYSIS

Abstract The safety of complex industrial processes such as chemical or power processes is of prime interest. In this paper, the monitorability indexes are introduced to build a monitoring instrumentation architecture assuring a predefined level of safety. Easily calculated from the bond graph model, these indexes give a measurement on how much the instrumentation system can be efficient in spite of a decrease in the performance of the process information system related to some faulty sensors. An example illustrates our approach.

Sensor Placement for Diagnosability Analysis in Thermofluid Processes

system in process engineerings safety is based on a monitoring system characterized by its ability to detect and to isolate faults. Considering single faults diagnosis, the existent monitoring systems treat only one category of additive faults at a time (actuator faults or sensor faults). The present paper considers thermofluid processes; the main objective is to identify the category of components (hydraulic actuators, thermal actuators, hydraulic sensors and thermal sensors) to which belongs the faulty element. Then one can seek the fault in the right category thanks to the existing methods. The innovative interest is thus to be able to isolate two component faults of different categories. On a second time, the set of thermal sensors is considered and an important result on their diagnosability is deduced. An application to an interconnected network of components in a thermofluid installation illustrates our results.