Yassine Koubaa

Simultaneous estimation of speed and rotor resistance in sensorless ISFOC induction motor drive based on MRAS Scheme

Usually the estimation of speed is achieved by assuming that the rotor resistance is constant throughout the operating range. In practice, the variation of this resistance depends on the temperature inside the machine. This paper proposes a simultaneous rotor speed and rotor resistance estimation method for a conventional indirect stator flux oriented controlled (ISFOC) induction motor drive. In order to estimate the both parameters, an adaptation algorithm based on the model reference adaptive system (MRAS) scheme for tuning the rotor speed and the rotor resistance is proposed. The reference and adjustable models, developed in stationary stator reference frame, are used in the MRAS scheme to estimate rotor speed and rotor resistance from measured terminal voltages and currents. The Integral-Proportional (IP) speed controller and Proportional-Integral (PI) current controller gains are calculated and tuned at each sampling time according to the new simultaneous estimation. A 3-phase induction motor has been used to verify the accuracy and feasibility of the proposed method. Simulation results show that the proposed method gives accurate estimations of simultaneous rotor speed and rotor resistance for a reference speed of the induction motor drive with nominal load torque is applied.

MRAS state estimator for speed sensorless ISFOC induction motor drives with Luenberger load torque estimation

This paper presents a novel method for estimating the load torque of a sensorless indirect stator flux oriented controlled (ISFOC) induction motor drive based on the model reference adaptive system (MRAS) scheme. As a matter of fact, this method is meant to inter-connect a speed estimator with the load torque observer. For this purpose, a MRAS has been applied to estimate the rotor speed with tuned load torque in order to obtain a high performance ISFOC induction motor drive. The reference and adjustable models, developed in the stationary stator reference frame, are used in the MRAS scheme in an attempt to estimate the speed of the measured terminal voltages and currents. The load torque is estimated by means of a Luenberger observer defined throughout the mechanical equation. Every observer state matrix depends on the mechanical characteristics of the machine taking into account the vicious friction coefficient and inertia moment. Accordingly, some simulation results are presented to validate the proposed method and to highlight the influence of the variation of the inertia moment and the friction coefficient on the speed and the estimated load torque. The experimental results, concerning to the sensorless speed with a load torque estimation, are elaborated in order to validate the effectiveness of the proposed method. The complete sensorless ISFOC with load torque estimation is successfully implemented in real time using a digital signal processor board DSpace DS1104 for a laboratory 3 kW induction motor.

Comparative study of different MPPT techniques for a stand-alone PV system

The (P-V) photovoltaic panels characteristic is nonlinear and highly depended on solar irradiation and temperature cell variations. For that reason, a Maximum Power Point Tracking (MPPT) algorithm is essential so that it is possible to draw peak power from the photovoltaic panel to obtain the maximized produced energy. The main purpose of MPPT technique is to make sure that the solar panel is producing the maximum power. This extracts the maximum amount of power at any given time. To figure out which MPPT technique has the best performances, this paper introduces six different MPPT algorithms existing in literature. In particular, these techniques are tested while the temperature cell was randomly varied. The simulation results are evaluated using Matlab-Simulink.

A Luenberger state observer for simultaneous estimation of speed and stator resistance in sensorless IRFOC induction motor drives

This paper characterizes the problem of implementing a simultaneous estimation of the stator resistance and the rotor speed of a sensorless indirect rotor flux oriented control (IRFOC) induction motor (IM) drive. For this purpose, the Luenberger state observer (LSO) is taken as a rudimentary method to simultaneously estimate the rotor speed and the stator resistance. Likewise, we suggest an adaptation algorithm related to the Lyapunov stability hypothesis to estimate the stator resistance and the rotor speed. The latter utilizes the estimated and measured stator currents and the estimated rotor flux. Along these lines, the control method that we suggest is liable to achieve a good performance after reducing the computational complexity through the use of the analytical relation in order to determine the Luenberger observer (LO) gain matrix. Additionally, compared to the previous observers, our observer is characterized by its simplicity, robustness as well as its ability to be implemented online. This article implies also the use of a typical PI regulator with feed-forward reparation expressions in the synchronous frame. But to regulate the rotor speed, we use the IP controller because of its effectiveness. Actually, its validity and efficiency are verified by the simulation results.

Application of statistical neuronal networks for diagnostics of induction machine rotor faults

Induction machines are extensively used in industries and are subject to unexpected breakdowns. It is necessary, therefore, to prevent them from such breakdown through the maintenance that works according to a well-trained planning. A considerable number of diagnosis techniques have been used such as Motor Current Signature Analysis (MCSA), Axial Flux Monitoring and Vibration Monitoring. This paper shows the effectiveness of the artificial neuronal network (radial basis function neuronal network and the probabilistic neuronal network) basis on MCSA for rotor faults diagnosis.

High gain observer design for a class of MIMO non uniformly observable uncertain systems

A high gain observer is proposed for a class of MIMO non uniformly observable systems including uncertainties. The gain of the proposed observer is issued from the resolution of a Lyapunov differential equation and its tuning can be achieved through the choice of a single scalar parameter. The observer design is made under an appropriate set of conditions given in terms of appropriate excitation. It is shown that in the absence of uncertainties, the observation error converges exponentially to zero. In the case where the uncertainties are bounded, this error can be made as small as desired by choosing high values of the observer design parameter. Simulation results are given for illustration purposes.

Unknown inputs observers for a class of non triangular nonlinear systems

In the present paper, we suggest a high gain observer for a class of multi-output nonlinear systems with nonlinearly parameterized unknown inputs. This observer should enable to estimate both the whole state and the unknown inputs, simultaneously. Not only does the gain of the observer not require the resolution of any dynamical system, but it is also explicitly given. For the sake of simplicity, the observer tuning is reduced to the choice of a single design parameter. The paper also reports on the simulation results for the sake of highlighting the performance of the proposed observer.