Zhenhai Li

Computation of a reference model for robust fault detection and isolation residual generation

This paper considers matrix inequality procedures to address the robust fault detection and isolation (FDI) problem for linear time-invariant systems subject to disturbances, faults, and polytopic or norm-bounded uncertainties.We propose a design procedure for an FDI filter that aims to minimize a weighted combination of the sensitivity of the residual signal to disturbances and modeling errors, and the deviation of the faults to residual dynamics from a fault to residual reference model, using the H∞-norm as a measure. A key step in our procedure is the design of an optimal fault reference model. We show that the optimal design requires the solution of a quadratic matrix inequality (QMI) optimization problem. Since the solution of the optimal problem is intractable, we propose a linearization technique to derive a numerically tractable suboptimal design procedure that requires the solution of a linear matrix inequality (LMI) optimization. A jet engine example is employed to demonstrate the effectiveness of the proposed approach.

Observer-based fault detection and isolation filter design for linear time-invariant systems

In this paper we consider a model-based fault detection and isolation problem for linear time-invariant dynamic systems subject to faults and disturbances. We use a state observer scheme that cancels the system dynamics and defines a residual vector signal that is sensitive only to faults and disturbances. We then design a stable fault detection and isolation filter such that the ℋ∞-norm of the transfer matrix function from disturbances to the residual is minimised (for fault detection) subject to the constraint that the transfer matrix function from faults to residual is equal to a pre-assigned diagonal transfer matrix (for isolation of possibly simultaneous occurring faults). Our solution is given in the form of linear matrix inequalities using state-space techniques, as well as a model matching problem using matrix factorisation techniques. A numerical example is given to illustrate the efficiency of the fault detection and isolation filter.

State Space Solution to the H - /H ∞ Fault Detection Problem

In this paper we give a state space solution to the H_-/H_∞ fault detection (FD) problem for linear time invariant dynamic systems. An H_-/H_∞ FD filter minimizes the sensitivity of the residual signal to disturbances while maintaining a minimum level of sensitivity to faults. We provide a state space realization of the optimal filter in an observer form via the solution of linear matrix inequalities (LMIs). A numerical example is given to illustrate the algorithm

Fault isolation filter with linear matrix inequality solution to optimal decoupling

In this paper we consider a model-based fault detection and isolation problem for linear time-invariant dynamic systems subject to faults and disturbances. We use an observer scheme that cancels the system dynamics and defines a residual vector signal that is sensitive only to faults and disturbances. We then design a stable fault isolation filter such that the Hinfin-norm of the transfer matrix function from disturbances to the residual is minimized (for fault detection) subject to the constraint that the transfer matrix function from faults to residual is equal to a pre-assigned diagonal transfer matrix (for fault isolation). The optimization of disturbance decoupling is accomplished via the help of linear matrix inequalities. A numerical example is also presented to illustrate the algorithm

A QMI APPROACH TO THE ROBUST FAULT DETECTION AND ISOLATION PROBLEM

Abstract This paper investigates the robust fault detection and isolation (FDI) problem for uncertain linear time-invariant (LTI) systems. An FDI filter minimizes the sensitivity of the residual signal to disturbances and modeling errors subject to the constraint that the transfer matrix function from the faults to the residual is closed to a diagonal transfer matrix function (for fault isolation). A solution of the optimization problem is presented via the formulation of a quadratic matrix inequality (QMI). A jet engine example is employed to demonstrate the effectiveness of our results.

Output Selection with Fault Tolerance via Dynamic Controller Design

Input-Output selection/placement for control systems has been an attractive research topic in particular under fault-free conditions. In this paper we present a methodology of output selection in a closed-loop framework with a view of fault tolerance capability. The principles with regards to the selection of sensors are reduced hardware redundancy, reduced costs and easier implementation, and acceptable degraded performance when faults occur. The selection of sensors is based upon both closed-loop control and fault tolerance objectives by solving an H1 optimization problem for each group of sensors sets via Linear Matrix Inequalities (LMIs). The proposed scheme is applied to a practical example of ride quality improvement of a high speed rail vehicle.

Output selection under control and fault detectability considerations

In a variety of practical engineering systems, i.e. aerospace, mechanical systems, railway vehicle systems, for a given requirement the range of possible locations for sensors is usually known, with the practical engineering issue of optimizing their location. Input-output selection/placement for control systems has been widely researched in particular under fault-free conditions. In this paper we discuss on the feasibility of an (output) sensor selection scheme in a closed-loop framework based on both control performance and fault detectability metrics. The selection of sensors is based upon both closed-loop control and fault detection objectives by solving a mixed H_/Hinfin optimization problem for each group of sensors available via Linear Matrix Inequalities (LMI). The efficacy of the scheme is illustrated via a numerical example.

Integrated design of dynamic controller with fault diagnosis and tolerance

Fault detection capability tends to become an integral part of control system design procedures for practical engineering systems. It is thus desirable fault diagnosis/tolerance functions to also be included in the controller design. In this context, we develop a generic observer-based feedback controller where the observer-part can also generate a residual signal for fault detection purposes. The design objectives is a mixture of H1 control and H1 fault detection and isolation. This multi-objective optimization problem is then formulated using Bilinear Matrix Inequalities (BMI) and a sub-optimal solution is achieved via transformation to Linear Matrix Inequalities (LMI). The developed approach and algorithm are verified in study of an application to a railway suspension system of ride quality maintenance.

A QMI Approach to the Robust Fault Detection and Isolation Problem

: This paper investigates the robust fault detection and isolation (FDI) problem for uncertain linear time-invariant (LTI) systems. An FDI filter minimizes the sensitivity of the residual signal to disturbances and modeling errors subject to the constraint that the transfer matrix function from the faults to the residual is closed to a diagonal transfer matrix function (for fault isolation). A solution of the optimization problem is presented via the formulation of a quadratic matrix inequality (QMI). A jet engine example is employed to demonstrate the effectiveness of our results.

A REFERENCE MODEL BASED ROBUST ℋ∞ FILTERING APPROACH TO FAULT DETECTION IN UNCERTAIN SYSTEMS

Abstract Model-based robust fault diagnosis aims to attenuate influence from model uncertainties on the residual while maintaining fault detection and isolation performance. In this paper, we consider robust residual generation for integral quadratic constrained (IQC) uncertain systems. The design consists of two steps. A reference model, incorporated into a robust ℋ∞ filtering framework, is set up to represent desired detection performance such as disturbance attenuation. Then, the extended robust ℋ∞ filtering problem is solved by constructing a nonlinear matrix inequality (NLMI). Linearization of the NLMI results in an tractable LMI solution, where an illustrative example follows to verify the algorithm.