Fatiha Nejjari

Passive Robust Fault Detection of Dynamic Processes Using Interval Models

Model-based fault detection relies on the use of a model to check the consistency between the predicted and the measured (or observed) behavior of a system. However, there is always some mismatch between the modeled and the real process behavior. Then, any model-based fault detection algorithm should be robust against modeling errors. One possible approach to take into account modeling uncertainty is to include all the uncertainty in system parameters using an interval model that allows generating an adaptive threshold. In this paper, the use of interval models in robust fault detection considering three schemes (simulation, prediction, or observation) is presented and discussed. The main contribution is to present a comparative study that allows identifying the benefits and drawbacks of using each scheme. This study will provide a guideline for the use of the proposed fault detection schemes in real applications. Finally, an intelligent servoactuator, proposed as a benchmark in the context of European Research Training Network DAMADICS, is used to illustrate the application and the comparative study of these interval-based fault detection schemes.

A GMDH neural network-based approach to passive robust fault detection using a constraint satisfaction backward test

This paper proposes a new passive robust fault detection scheme using non-linear models that include parameter uncertainty. The non-linear model considered here is described by a group method of data handling (GMDH) neural network. The problem of passive robust fault detection using models including parameter uncertainty has been mainly addressed by checking if the measured behaviour is inside the region of possible behaviours based on the so-called forward test since it bounds the direct image of an interval function. The main contribution of this paper is to propose a new backward test, based on the inverse image of an interval function, that allows checking if there exists a parameter in the uncertain parameter set that is consistent with the measured system behaviour. This test is implemented using interval constraint satisfaction algorithms which can perform efficiently in deciding if the measured system state is consistent with the GMDH model and its associated uncertainty. Finally, this approach is tested on the servoactuator being a FDI benchmark in the European Project DAMADICS.

Leak Localization in Water Networks: A Model-Based Methodology Using Pressure Sensors Applied to a Real Network in Barcelona [Applications of Control]

The efficient distribution of water is a subject of major concern for water utilities and authorities [1]. While some leaks in water distribution networks (WDNs) are unavoidable, one of the main challenges in improving the efficiency of drinking water networks is to minimize leaks. Leaks can cause significant economic losses in fluid transportation and extra costs for the final consumer due to the waste of energy and chemicals in water treatment plants. Leaks may also damage infrastructure and cause third-party damage and health risks. In many WDNs, losses due to leakage are estimated to account up to 30% of the total amount of extracted water [2]; a very important issue in a world struggling to satisfy water demands of a growing population.

Robust quasi-LPV model reference FTC of a quadrotor UAV subject to actuator faults

Abstract A solution for fault tolerant control (FTC) of a quadrotor unmanned aerial vehicle (UAV) is proposed. It relies on model reference-based control, where a reference model generates the desired trajectory. Depending on the type of reference model used for generating the reference trajectory, and on the assumptions about the availability and uncertainty of fault estimation, different error models are obtained. These error models are suitable for passive FTC, active FTC and hybrid FTC, the latter being able to merge the benefits of active and passive FTC while reducing their respective drawbacks. The controller is generated using results from the robust linear parameter varying (LPV) polytopic framework, where the vector of varying parameters is used to schedule between uncertain linear time invariant (LTI) systems. The design procedure relies on solving a set of linear matrix inequalities (LMIs) in order to achieve regional pole placement and H∞ norm bounding constraints. Simulation results are used to compare the different FTC strategies.

An Interval NLPV Parity Equations Approach for Fault Detection and Isolation of a Wind Farm

In this paper, the problem of fault diagnosis of a wind farm is addressed using interval nonlinear parameter-varying (NLPV) parity equations. Fault detection is based on the use of parity equations assuming unknown but bounded description of the noise and modeling errors. The fault detection test is based on checking the consistency between the measurements and the model, by finding if the formers are inside the interval prediction bounds. The fault isolation algorithm is based on analyzing the observed fault signatures online and matching them with the theoretical ones obtained using structural analysis. Finally, the proposed approach is tested using the wind farm benchmark proposed in the context of the wind farm fault-detection-and-isolation/fault-tolerant-control competition.

Sensor placement for leak detection and location in water distribution networks

The performance of a leak detection and location algorithm depends on the set of measurements that are available in the network. This works presents and optimization strategy that maximizes the leak diagnosability performance of the network. The goal is to characterize and determine a sensor configuration that guarantees a maximum degree of disnosability while the sensor configuration cost satifies a budgetary constraint. To efficiently handle the complexity of the distribuion networl an efficient branch and bound search strategy based on a strucutrual model is used. However, in order to reduce even more the size and the complexity of the problem the present work proposes to combine this methodology with clustering techniques. The strategy developed in this work is successfully applied to determine the optimal set of pressure sensors that should be installed to a District Metered Area in the Baarcelona Water Distribution Network.

Fault tolerant control of the wind turbine benchmark using virtual sensors/actuators

Abstract In this paper, the problem of Fault Tolerant Control (FTC) of the wind turbine benchmark is addressed. The paper proposes the use of virtual sensor/actuator approaches to deal with sensor and actuator faults, respectively. The paper suggests the reformulation of these FTC schemes that have been already proposed in state space form in input/output form. The FTC module will use the information from the Fault Detection and Isolation (FDI) module previously designed using set-membership techniques. A fault estimation scheme is also proposed based on batch least squares approach. The performance of the proposed FTC schemes will be assessed using the proposed fault scenarios considered in the FTC benchmark.

A Fault-Hiding Approach for the Switching Quasi-LPV Fault-Tolerant Control of a Four-Wheeled Omnidirectional Mobile Robot

This paper proposes a reference model approach for the trajectory tracking of a four-wheeled omnidirectional mobile robot. In particular, the error model is brought to a quasi-linear-parameter-varying (LPV) form suitable for designing an error-feedback controller. It is shown that, if polytopic techniques are used to reduce the number of constraints from infinite to finite, a solution within the standard LPV framework could not exist due to a singularity that appears in the possible values of the input matrix. Adding a switching component to the controller allows solving this problem. Moreover, a switching LPV virtual actuator is added to the control loop in order to obtain fault tolerance within the fault-hiding paradigm, keeping the stability and some desired performances under the effect of actuator faults without the need of retuning the nominal controller. The effectiveness of the proposed approach is shown and proved through simulation and experimental results.

Fault Diagnosis Based on Causal Computations

This paper focuses on residual generation for model-based fault diagnosis. Specifically, a methodology to derive residual generators when nonlinear equations are present in the model is developed. A main result is the characterization of computation sequences that are particularly easy to implement as residual generators and that take causal information into account. An efficient algorithm, based on the model structure only, which finds all such computation sequences, is derived. Furthermore, fault detectability and isolability performances depend on the sensor configuration. Therefore, another contribution is an algorithm, also based on the model structure, that places sensors with respect to the class of residual generators that take causal information into account. The algorithms are evaluated on a complex highly nonlinear model of a fuel cell stack system. A number of residual generators that are, by construction, easy to implement are computed and provide full diagnosability performance predicted by the model.

Robust state-feedback control of uncertain LPV systems : an LMI-based approach

Abstract In this paper, the problem of designing an LPV state-feedback controller for uncertain LPV systems that can guarantee some desired bounds on the H ∞ and the H 2 performances and that satisfies some desired constraints on the closed-loop poles location is considered. In the proposed approach, the vector of varying parameters is used to schedule between uncertain LTI systems. The resulting idea consists in using a double-layer polytopic description so as to take into account both the variability due to the parameter vector and the uncertainty. The first polytopic layer manages the varying parameter and is used to obtain the vertex uncertain systems, where the vertex controllers are designed. The second polytopic layer is built at each vertex system so as to take into account the model uncertainties and add robustness into the design step. Under some assumptions, the problem reduces to finding a solution to a finite number of LMIs, a problem for which efficient solvers are available nowadays. The solution to the multiobjective design problem is found both in the case when a single fixed Lyapunov function is used and when multiple parameter-varying Lyapunov functions are used. The validity and performance of the theoretical results are demonstrated through a numerical example.