Eric Blanco

Coupling Pattern Recognition With State Estimation Using Kalman Filter for Fault Diagnosis

This paper deals with a diagnosis tool based on a pattern recognition approach associated with Kalman interpolator/extrapolator. The first aim is to decrease the number of measurements to realize while increasing the learning database contents using a Kalman state estimator. The second one is to estimate, from the initial set of measured data, future states of the studied process. A 5.5-kW induction motor bench is used as an application to validate this approach. First, a signature is determined in order to monitor the different operating modes evolution. Diagnostic features are extracted only from current and voltage sensors. Then, a feature selection method is applied in order to select the most relevant features for diagnosis. Finally, a Kalman filter algorithm is developed in order to interpolate the known states and to predict evolution toward new ones. A new diagnosis tool is then designed handling continuous evolution (severity, load) inside the different operating modes (healthy, stator fault, ...).

Fault Detection and Diagnosis in a Set “Inverter–Induction Machine” Through Multidimensional Membership Function and Pattern Recognition

Nowadays, electrical drives generally associate inverter and induction machine. Thus, these two elements must be taken into account in order to provide a relevant diagnosis of these electrical systems. In this context, the paper presents a diagnosis method based on a multidimensional function and pattern recognition (PR). Traditional formalism of the PR method has been extended with some improvements such as the automatic choice of the feature space dimension or a ldquononexclusiverdquo decision rule based on the k-nearest neighbors. Thus, we introduce a new membership function, which takes into account the number of nearest neighbors as well as the distance from these neighbors with the sample to be classified. This approach is illustrated on a 5.5 kW inverter-fed asynchronous motor, in order to detect supply and motor faults. In this application, diagnostic features are only extracted from electrical measurements. Experimental results prove the efficiency of our diagnosis method.

Robust Filtering for Linear Time-Invariant Continuous Systems

The problem of robust filtering for linear time-invariant (LTI) continuous systems subject to parametric uncertainties is treated in this paper through transfer function and polynomial representations, and then in the state-space domain. The basic idea consists of introducing the gradient of the estimation error with respect to the uncertain parameters in the optimization scheme via a epsiv-contaminated model. The general solution to the problem is given in the transfer function representation while, in the polynomial framework, the causal estimator is obtained by means of a spectral factorization and a Diophantine equation. The state-space realization of the causal estimator is discussed. Examples show the ability of the proposed technique to provide a reliable estimation in presence of model uncertainty.

A clock network of distributed ADPLLs using an asymmetric comparison strategy

In this paper, we describe an architecture of a distributed ADPLL (All Digitall Phase Lock Loop) network based on bang-bang phase detectors that are interconnected asymmetrically. It allows an automatic selection between two operating modes (uni- and bidirectional) to avoid mode-locking phenomenon, to accelerate the network convergence and to improve the robustness to possible network failures in comparison to simple unidirectional mode.

The

The Hinfin smoothing problem for continuous systems is treated in a state space representation by means of variational calculus techniques. The smoothing problem is introduced in an Hinfin criterion by means of an artificial discontinuity that splits the problem in term of Hinfin forward and Hinfin backward filtering problems. Hence, the smoother design is realized in three steps. First, a forward filter is developed. Secondly, a backward filter is developed taking into account the backward Markovian model. The third step consists of combining the two previous steps in order to compute the Hinfin smoothed estimate. An example shows the efficiency of this proposed smoother

H_infty

The robust filter design and the robust feedforward controller design are particular cases of a larger class of problems: the robust open-loop problems. In this article, we consider a class of uncertain open-loop plants, where a filter needs to be designed to ensure that the plant satisfies chosen specifications. The representation of uncertainties is made in a very general framework: the Linear Fractional Transformation (LFT). Associated with the Dynamic Integral Quadratic Constraints framework, it allows the consideration of many classes of structured uncertainties. This paper proves that the design of a filter ensuring a robust L2-gain or H2 performance for the complete plant can be expressed as a convex optimization problem involving Linear Matrix Inequalities Constraints which can be solved using an efficient algorithm.

Fixed-Interval Smoothing Problem for Continuous Systems

The robust L2-gain estimation is investigated for general uncertain systems with structured uncertainties. A new estimation structure is introduced: the Augmented-Gain Observer which encompasses both filters and observers and allows robust estimation even for some classes of unstable systems. Our approach is based on a separation of graphs theorem using frequency dependent Integral Quadratic Constraints. We prove that the design of an Augmented-Gain Observer ensuring a robust L2-gain performance can be expressed as a convex optimization problem. This problem involves Linear Matrix Inequalities constraints and can be solved using an efficient algorithm. A numerical example illustrates the interest of the method.

An LMI solution for a class of robust open-loop problems

This paper deals with the tracking and the prediction of the evolution of the system operation. The aim is to define a forecast of future operating state of the process by using the previous state. First of all, a signature is determined in order to monitor the evolution of different operating modes. For this purpose, on the example of an induction machine, diagnostic features are extracted from current and voltage measurements without any other sensors. Then, a feature selection method is applied in order to select the most relevant features which define the representation space. A polynomial approach of tracking evolution is presented. Next, a Kalman algorithm is developed to predict evolution and to allow pre-empting on the appearance of a fault and the accelerated ageing of system. Finally these two approaches are applied and compared with an induction machine of 5.5 kW with squirrel-cage.

Robust L 2 -Gain Observation for structured uncertainties: An LMI approach

An H/sub /spl infin// smoother has been developed (Blanco et al., 2000) and it gives good results for noise uncertainties. Nevertheless, when uncertain parameters appear, its performance decreases significantly. Therefore, an estimator robust to noise uncertain properties and parameter uncertainties is presented. The robust H/sub /spl infin// smoother for uncertain systems is developed as a combination of two robust H/sub /spl infin// filters. The robust performance, for both noise and parameter uncertainties, of this new approach is evaluated on a simple example.

Beyond the diagnosis: the forecast of state system Application in an induction machine

The robust estimation problem for uncertain discrete-time systems is treated in this paper considering parametric uncertainties in the polynomial framework. The estimator is the minimiser of a cost function written in a ∈-contaminated form. Uncertainties are not formally modelled. It is shown that, in the present context, the proposed approach can be connected to stochastic modelling of uncertainties. The optimal robust estimator is obtained by computing a spectral factorisation and by solving a single Diophantine equation. An example shows the efficiency of the proposed method.