94 publications from this institution
du domaine d'dtude.1.
Experimental results on the magnetic saturation effects on the control of a small synchronous reluctance machine (600 W) are presented in this paper. A new model of the machine, including saturation effects and cross magnetization is first developed. An approach based on total and mutual inductances is followed instead of the traditional approach through magnetizing and leakage inductances. All of the electrical and mechanical parameters measurements required for the simulations are given. A rotor-oriented vector control of the synchronous reluctance machine is achieved with a digital-signal-processor (DSP) board (TMS320C31) and experimental results are presented. When the magnetic circuit is saturated, the simulation with the developed model shows good, accurate results when compared to the experimental ones.
This paper describes an identification procedure by output error to estimate the parameters of an induction motor. This approach requires an optimisation algorithm. Two different strategies are considered: a genetic algorithm and a quasi-Newton algorithm. The last one requires the evaluation of the gradient and the Hessian in order to minimise a criterion in opposition to the first one. These two techniques have the advantage to be applied to linear and nonlinear models. They can take into account some constraints such as knowledge on electrical parameters. The experimental results prove the effectiveness of these two algorithms.
In the present paper, the equivalence between the magnetic images method and the classical variable separation method is set up. These methods are separately applied to calculate the magnetic scalar potentials and the magnetic vector potentials in the air-gap of some cylindrical synchronous machines, excited by surface permanent magnets with radial and/or azimuthal magnetisation, and without polar pieces. The rotor and the stator are assumed slotless, and the whole electric machine non saturated. The expressions of the magnetic scalar potentials of each method are interpreted, and their equivalence is established. The Mayn purpose of this paper is then the demonstration that the magnetic images method is applicable for the study of simplified synchronous machines. The interest of the magnetic images use in comparison with the variable separation method is shown.
The major drawback of usual dual-stator ac machines, when supplied by a voltage-source inverter (VSI), is the occurrence of extra harmonic currents. These extra currents circulate only in the stator windings and cause additional losses. This paper deals with the modeling and design of dual-stator winding ac machines for safe operation with a VSI. A new reference frame model for a dual-stator induction machine (DSIM), including mutual leakage coupling, is proposed. This model allows us to highlight the previously mentioned circulating harmonic currents. The leakage inductance associated with these harmonics is shown to have quite a small value, highly depending on the coil pitch. It is also shown that full pitch is required, and that special slot shape designs should be investigated, to limit the magnitude of circulating currents. Experimental results on a prototype of a DSIM are presented and they show a very good correlation with theoretical curves.
In electrical engineering, static converters generate harmonics; consequently the losses in the conductors increase. To use High-Tc superconductors for current transportation, it is important to evaluate and to measure their losses. Under the assumption that the superconductor is only exposed to its self-field, Bean's model is used to compute these losses. This work deals with the losses in a cylindrical High-Tc superconductor current-lead. It is made of compacted composite of Bi-Pb-Sr-Ca-Cu-O-2223, the value of its critical current is about 100 A (measured with I /spl mu/V/cm criteria). Two types of measuring methods are used. The first of them uses a differential amplifier associated with a computer, which calculate the instantaneous power by numerical integration. The second series of experiments, valid only with sinusoidal current, are made using a lock-in amplifier measuring the R.M.S. voltage. The value of transport current is high enough to be compatible with applications. Both methods give nearly the same results for sinusoidal signal. Measurements are made with non-sinusoidal current, to do that the superconductor sample is fed by a dimmer switch. The experiments show that the losses are essentially a function of the maximum value of the current, not of the firing angle. We conclude that the shape of the current is not important compared to its peak value.
A new type of large torque air-cored cryogenic ac machine has been presented earlier. The experimental results obtained with the first prototype Supersat 001 are discussed. The experimental results are compared with the theorical ones deduced from a three-dimensional analytical model. The main results concern the static and dynamic behavior and the analysis of the losses which have led to the consideration of a new design of the first prototype.
This paper is intended to show how to replace a fuzzy controller by a neural network one in a rotor field oriented control for induction motors. The output surface of the fuzzy controller is approached by a neural network. The neural network structure used for this controller is presented as well as the learning process necessary to prepare the neural network controller to act like a fuzzy one.Experimental results on position control and speed control under load are presented and commented as well as a discussion and comparison with classical Pl controllers.
This paper deals with a new kind of superconducting limiter. Usually, superconducting current limiters are based on the transition from superconducting state to the normal state by overtaking the critical current. This new superconducting current limiter has no current transport in the superconductor. Indeed, in our structure, the limitation is based on the saturation of a high temperature bulk superconductor by the magnetic flux. We show that our structure enables one to limit the current.
This paper deals with a new kind of superconducting limiter which is not based on the transition from the superconducting state to the normal one by overtaking the critical current. This new superconducting current limiter has no current transport in the superconductor. Indeed, in our structure, the current limitation is based on the saturation of a high temperature bulk superconductor by the magnetic flux. The principle of this new type of limiter has been explained in a previous paper. Now we present the design of this kind of limiter and we show the influence of the temperature on its functioning.
ABSTRACT In this paper we propose a model of the synchronous machine for fast digital computation of its transient behaviour. This modelling avoids numerical inversion of the inductance matrix even when some stator phases are open-circuited and allows then a large saving in the computation time. The obtained models are tested in the simulation of some sudden short circuits of an alternator. Simulation and experimental results are compared as well as the computation time using this modelling and the usual one.
The major drawback of usual dual-stator AC machines supplied by a voltage source inverter (VSI) is the occurrence of large circulating harmonic currents, causing extra losses. This paper deals with the design of dual-stator winding AC machines for safe operation with VSI. A new reference frame model of a dual-stator induction machine (DSIM), including mutual leakage coupling, is proposed. The leakage inductance associated with the inaccurate harmonics is shown to have a quite small value, highly depending on the coil pitch. It is also shown that full pitch is required, and that special designs of slot shape should be investigated to limit the negative impact of circulating currents. Experimental results on a prototype of DSIM are presented and they show very good correlation with the theoretical curves.
Induction motor plays a very important role in industry applications. Early fault detection is one of the main problems. So, in order to detect an incipient defect, we must pay a special attention to the spectral analysis of instantaneous power. The effectiveness of this method is based on the reliability of the signals. The study of the product of voltage and current allows the detection of a rotor fault in induction motor. Several experiments were made on a test bed of 3 kW. A spectral analysis of instantaneous power for the detection of a half broken rotor bar is used to provide an accurate diagnosis. Comparisons between spectrums database when the motor is healthy and under rotor faults are presented and commented. The results obtained show the possibility to detect a half broken rotor bar in induction motor.
