Michel Verge

Tactile objects based on an amplitude disturbed diffraction pattern method

Tactile sensing is becoming widely used in human-computer interfaces. Recent advances in acoustic approaches demonstrated the possibilities to transform ordinary solid objects into interactive interfaces. This letter proposes a static finger contact localization process using an amplitude disturbed diffraction pattern method. The localization method is based on the following physical phenomenon: a finger contact modifies the energy distribution of acoustic wave in a solid; these variations depend on the wave frequency and the contact position. The presented method first consists of exciting the object with an acoustic signal with plural frequency components. In a second step, a measured acoustic signal is compared with prerecorded values to deduce the contact position. This position is then used for human-machine interaction (e.g., finger tracking on computer screen). The selection of excitation signals is discussed and a frequency choice criterion based on contrast value is proposed. Tests on a sandwich plate (liquid crystal display screen) prove the simplicity and easiness to apply the process in various solids.

Active damage detection and localization applied to a composite structure using piezoceramic patches

This paper describes the application of an active structural health monitoring technique for a composite plate bonded with a distributed piezoceramic patches array. To maximize the performance of a structural health monitoring (SHM) system, an optimal placement of these patches has been proposed using grammians of controllability and observability and the Hœ modal norm. The proposed approach was established via finite element modelization (FE) of the composite structure. With this optimal placement, an active structural health monitoring scheme has been accomplished. Therefore, using spatial information given by the distributed sensors, we have proposed two damage indices (DI). These indices are able to detect and localize damages. The first DI is based upon non-parametric frequency response function estimates; the damage is detectable when changes in the estimates exceed their normal statistical bounds. The second DI uses change in the energy signal sensor to localize damage existing in the composite structure. The efficiency of the proposed approach is demonstrated through experimental tests.

Proper orthogonal decomposition applied to structural health monitoring

This paper describes the application of proper orthogonal decomposition (POD) method in the context of structural health monitoring of a smart structure, bonded with a distributed piezoceramic patches array. Using measurements given by the distributed sensors, we propose a damage index (DI) based on change in angle between subspaces. Furthermore, from statistics, we propose a threshold that automatically decides if the structure is in healthy or damaged state. The efficiency of the proposed approach is validated on composite plate with two different damages (impact damage and additive mass).

FRICTION IDENTIFICATION WITH GENETIC ALGORITHMS

Abstract Identification of the function part of the analytical model of the movement of a mass along vertical guide way is made. Indeed, knowledge of friction parameters is necessary to detect wear or mechanical failure. Because of the non linearity the identification is made using Genetic Algorithms (GA). After many simulations, the proposed approach is tested with the recorded experimental data from a test bed.

Faults Detection by Marginalized Particle Filters: Application to a Drilling Process

Abstract Real-time faults diagnosis in processes is an essential step to improve their efficiencies. In this paper, by using Marginalized Particle Filters (MPF) for parameter identification, we show the possibility of its exploitation to achieve a Faults Detection and Isolation (FDI) scheme. The idea behind the proposed methodology is motivated by the assumption based on the hypothesis that occurrence of a well-defined fault can bring a process in a well-defined domain. As a result, the whole process model is developed so as to get a part which stands for the faulty case and another that represents the safety one. Thereby, the distribution of these two process states is updated through a Kalman filter and the concerned state is involved by using a particle filter. Experiments and detailed performances analysis using both simulated and real data, obtained from a drilling process used in oilfield industry, indicate that the proposed approach is efficient in terms of tracking ability of parameters change and consequently for FDI task.

New element for sandwich beams with piezoelectric layers

A new element is developed for the analysis of laminated/sandwich beams incorporating piezoelectric layers. Beam theory employed here satisfies the interface stress and displacement continuity and has zero shear stress on the top and bottom surfaces of the beam. Piezoelectric layers can be placed arbitrarily and can act as sensors and actuators. The transverse shear deformation in the form of trigonometric sine function is incorporated to define the transverse shear strain. The displacement and stress variations along the length and thickness obtained are presented.

Stochastic adaptive learning rate in an identification method: An approach for on-line drilling processes monitoring

On-line drilling processes monitoring is an essential task in enhancing their performances. In oilfield industry, dysfunctions that might occur have to be detected at the earliest possible stage in order to preserve drilling efficiency. This paper deals with a methodology for drilling processes monitoring by identifying time varying parameters. The basic idea behind the proposed algorithm is to improve the tracking ability of parameters change by means of an identification method using a new approach to adjust the forgetting factor. The effectiveness of the developed method is highlighted through experimental data obtained from tests campaign.

Structural Damage Diagnosis Using Subspace Based Residual and Artificial Neural Networks

Abstract In this paper, an Artificial Neural Network (ANN) based approach using a new non-parametric residual is presented for damage diagnosis. The residual is associated with Observability null-space of the system and is generated by using a parity matrix obtained from covariance driven output-only Subspace Identification (SubID) algorithm. Training of ANN is established using residuals generated from Finite Element (FE) model of the structure under consideration for different defect cases. This trained ANN is in turn used to identify the predefined defect types, in semi-real time, on the actual structure. The proposed methodology is applied to an active composite beam structure and its effectiveness to identify damage is studied by testing a trained ANN with data obtained from both simulation and experimentation. Promising results were obtained showing that, it was possible to distinguish between healthy and damaged states with good accuracy and repeatability.

Recursive modal parameter estimation using output-only subspace identification for structural health monitoring

Precise identification of structural properties is a vital step towards detection and localization of damage in structures. In this paper the problems of improving accuracy of modal parameter estimates and automatic elimination of spurious modes in covariance driven output-only subspace identification method are addressed. An iterative procedure is proposed, which in the first step, actively modifies the excitation signal, resulting in improvements in identification results. In the second step, for spurious mode elimination, an alternative stabilization histogram is introduced to automatically combine and extract identified modal parameters. It is shown that use of measured output signals, of different sampling rates, along with combination of identified results on a single stabilization histogram, can enhance the effectiveness of spurious modes rejection. One numerical and two experimental examples, on the modal parameter estimation of a composite beam and an aluminum plate (CACTUS) are presented to demonstrate the efficacy of the iterative procedure. The proposed algorithm allows automatic and accurate generation of modal parameter residuals for structural health monitoring applications.

Model reduction for active vibration control

In active vibration control, the modeling of the mechanical structure is generally done using the finite element method. The model obtained must be reduced to compute the controller. Model reduction must achieve two objectives: the reduced model must keep its initial dynamic behavior in the frequency range of interest; and the influence of neglected dynamics must be minimized to limit the risk of spillover. In this paper, we propose an original criterion for the selection of the controlled dynamics. This criterion has the great advantage to be independent of the placement of actuators and sensors. First, the structure is successively reduced to only one mode. For each one, it is obvious to find the optimal location of one actuator and one sensor. Thus, we obtain as many SISO systems of second order as the number of modes. Then, the Hinfinity norms of these systems are computed and the modes are sorted regarding to their influence in the global response of the system. To show the efficiency of the method, it is applied to an experimental structure: an LQR controller is designed and several tests are performed. A comparison is also done with other classical techniques of reduction.