Showing papers by "Ron J. Patton published in 2002"

Model-based fault diagnosis in dynamic systems using identification techniques

Abstract: 1. Introduction.- 2. Model-based Fault Diagnosis Techniques.- 3. System Identification for Fault Diagnosis.- 4. Residual Generation, Fault Diagnosis and Identification.- 5. Fault Diagnosis Application Studies.- 6. Concluding Remarks.- References.

Robust fault detection in descriptor linear systems via generalized unknown input observers

Abstract: A new parametric approach for robust fault detection in descriptor linear multivariable systems with unknown disturbances is proposed. The residual is generated using a fullorder generalized state observer. Based on a recently proposed parametric eigenstructure assignment approach, parameterizations of the observer gain and the eigenvectors of the observer system are presented. By combining the parameterizations of the observer eigenvectors and an established condition for disturbance decoupling in descriptor linear systems, the effect of the disturbance to the residual signal is decoupled. A simple algorithm is presented. An example shows the effect of the proposed approach.

SVD-based complexity reduction to TS fuzzy models

Abstract: One of the typical important criteria to be considered in real-time control applications is the computational complexity of the controllers, observers, and models applied. In this paper, a singular value decomposition (SVD)-based complexity reduction technique is proposed for Takagi Sugeno (TS) fuzzy models. The main motivation is that the TS fuzzy model has exponentially growing computational complexity with the improvement of its approximation property through, as usually practiced, increasing the density of antecedent terms. The reduction technique proposed here is capable of defining the contribution of each local linear model included in the TS fuzzy model, which serves to remove the weakly contributing ones as according to a given threshold. Reducing the number of models leads directly to the computational complexity reduction. This work also includes a number of numerical and application examples.

Polytopic and TS models are nowhere dense in the approximation model space

Abstract: We show in this paper that the set of functions, consisting of polytopic or TS models constructed from finite number of components, is nowhere dense in the approximation model space, if that is defined as a subset of continuous functions. This topological notion means that the given set of functions lies "almost discretely" in the space of approximated functions. As a consequence, by means of the mentioned models we cannot approximate in general continuous functions arbitrarily well, if the number of components are restricted. Thus, only functions satisfying certain conditions can be approximated by such models, or alternatively, we need unbounded number of components. The possible solutions are outlined in the paper.

Fault Diagnosis of an Electro-pneumatic Valve Actuator Using Neural Networks With Fuzzy Capabilities

Abstract: The early detection of faults (just beginning and still developing) can help avoid system shutdown, breakdown and even catastrophes involving human fatalities and material damage. Computational intelligence techniques are being investigated as an extension to the traditional fault diagnosis methods. This paper discusses the neuro-fuzzy approach to modelling and fault diagnosis, based on the TSK/Mamdani approaches. An application study of an electro-pneumatic valve actuator in a sugar factory is described. The key issues of finding a suitable structure for detecting and isolating ten realistic actuator faults are outlined.

Fault diagnosis of industrial actuators in a sugar factory: a European study

Abstract: The likelihood that actuator systems (e.g. control valves, servo motors, positioners) will malfunction is significant when these components are installed in harsh environments (e.g. with high temperature, humidity, pollution, chemical solvents, aggressive media etc.). Determination of small (incipient, hard to detect) faults before they become serious clearly has an important influence on the actuator's predicted lifetime.

Design methodology of tensor product based control models via HOSVD and LMIs

Abstract: This paper aims at solving the conflicts of the computational needs of building tensor product (TP) based control models having high approximation accuracy and practical aspects of their applicability w.r.t. system stability and feasibility. Therefore first we propose a HOSVD (higher order singular value decomposition) based methodology capable of implementing any differential equation systems of a dynamic model in TP model form with specific basis functions whereupon the linear matrix inequalities (LMI) based controller design techniques and stability analysis can be executed. Second, we intend to find a tradeoff between the TP modelling accuracy, hence system performance, and the controller complexity which is bounded by the available real time computation power at hand. As an example a detailed control design is given.

Automatic TS-model representation of analytic differential equations for Tanaka's LMI based control design

Abstract: We propose a HOSVD (Higher Order Singular Value Decomposition) based approach to LMI based control design and stability analysis techniques. The proposed method deals with the analytic differential equations of dynamic models and automatically finds their TS model representation. As a subsequent advantage, it optimizes the complexity of the resulted TS model with respect to approximation accuracy. As an example a detailed control design is given.