Mohammed Belkheiri

Different linearization control techniques for a quadrotor system

In this paper, we propose different linearization control algorithms to solve the stabilization problem of the quadrotor. First we introduce the nonlinear model of the quadrotor. Then using tangent linearization method, a linear model is generated of the system where decentralized and centralized LQR control methods are applied. The second strategy is based on exact feedback linearization of the nonlinear model of the quadrotor. The comparison between these methods is highlighted by simulations to show effectiveness of the proposed methods.

Feedback linearisation control of an induction machine augmented by single-hidden layer neural networks

We consider adaptive output feedback control methodology of highly uncertain nonlinear systems with both parametric uncertainties and unmodelled dynamics. The approach is also applicable to systems of unknown, but bounded dimension. However, the relative degree of the regulated output is assumed to be known. This new control strategy is proposed to address the tracking problem of an induction motor based on a modified field-oriented control method. The obtained controller is then augmented by an online neural network that serves as an approximator for the neglected dynamics and modelling errors. The network weight adaptation rule is derived from the Lyapunov stability analysis, that guarantees boundedness of all the error signals of the closed-loop system. Computer simulations of an output feedback controlled induction machine, augmented via single-hidden-layer neural networks, demonstrate the practical potential of the proposed control algorithm.

Neural network augmented backstepping control for an induction machine

A new control approach is proposed to address the tracking problem of an induction machine based on a modified Field-Oriented Control (FOC) method. In this approach, one relies first on a partially known model to the system to be controlled using a backstepping control strategy. The obtained controller is then augmented by an Adaptive Neural Network (NN) that serves as an approximator for the neglected dynamics and modelling errors. The proposed approach is systematic, and exploits the known non-linear dynamics to derive the stepwise virtual stabilising control laws. At the final step, an augmented Lyapunov function is introduced to derive the adaptation laws of the network weights. The effectiveness of the proposed controller is demonstrated through computer simulation.

Backstepping Control Augmented by Neural Networks For Robot Manipulators

A new control approach is proposed to address the tracking problem of robot manipulators. In this approach, one relies first on a partially known model to the system to be controlled using a backstepping control strategy. The obtained controller is then augmented by an online neural network that serves as an approximator for the neglected dynamics and modeling errors. The proposed approach is systematic, and exploits the known nonlinear dynamics to derive the stepwise virtual stabilizing control laws. At the final step, an augmented Lyapunov function is introduced to derive the adaptation laws of the network weights. The effectiveness of the proposed controller is demonstrated through computer simulation on PUMA 560 robot.

Control of three-level NPC inverter based grid connected PV system

This paper presents the control of a photovoltaic distributed generation system based on dual-stage topology of DC-DC boost converter and three-level neutral-point-clamped (NPC) voltage source inverter (VSI). Decoupling control strategy of three-level VSI is proposed to control the current injected into the grid, reactive power compensation, and DC-link voltage. The resulting system is able to extract the maximum power from photovoltaic generator, to achieve sinusoidal grid currents, and to ensure reactive power compensation. The voltage-balancing control of two split DC capacitors of the three-level VSI is achieved using three-level space vector modulation with balancing strategy based on the effective use of the redundant switching states of the inverter voltage vectors. The proposed system performance is investigated at different operating conditions.

Modeling flashover voltage (FOV) of polluted HV insulators using artificial neural networks (ANNs)

This paper attempts to apply artificial intelligent techniques in high voltage applications and especially to estimate the critical flashover voltage (FOV) for polluted insulators, using experimental measurements carried out in an insulator test station according to the IEC norm and a mathematical model based on the characteristics of the insulator: the diameter, the height, the creepage distance, the form factor and the equivalent salt deposit density and estimates the critical flashover voltage. Two types of artificial neural networks (ANNs) are designed to establish a nonlinear model between the above mentioned characteristics and the critical flashover voltage. The ANNs models, algorithms, and tools have been developed using the software package Matlab. The obtained results are promising and insure that artificial intelligent techniques can estimate the critical flashover voltage for new designed insulators with different operating conditions and constitute an indispensable models that can be used in field simulations of various parameters for polluted insulators. Further comparative analysis of the estimated results with the measured data collected from the site measurement amply demonstrate the effectiveness of the use of artificial intelligent techniques for modeling (ANNs) of FOV.

Adaptive nonlinear observer augmented by radial basis neural network for a nonlinear sensorless control of an induction machine

This paper presents adaptive neural network nonlinear observer associated with a sensor less nonlinear feedback linearisation controller for induction machine. The proposed observer is used to estimate the mechanical speed using the stator currents measurements and the supplied input voltages; whereas the load torque (unknown disturbance) is estimated using online radial basis neural network function approximation. The stability of the proposed controller-observer is achieved using Lyapunov function. Hence, simulation results have been performed under MATLAB/Simulink shows clearly the performance of the proposed algorithm.

Vehicle online parameter estimation using a nonlinear adaptive observer

This paper presents a nonlinear adaptive observer for the estimation of the wheel stiffness and radius. The estimated parameters will be used for controller synthesis and supervision for vehicle applications. The proposed adaptive observer uses the angular velocity as an input of the system, whereas the angular position and the vehicle velocity are considered as a measured state vector. The adaptive observer is designed based on a nonlinear model of the vehicle quarter model and the adaptive law of the parameters is derived using Lyapunov analysis. Simulation results show clearly the effectiveness of the proposed observer that achieved a good performance for vehicle parameter estimation.

FPGA based control of a PWM inverter by the third harmonic injection technique for maximizing DC bus utilization

The objective of this paper is to design and implement a third harmonic injection PWM strategy THIPWM on an FPGA, to control a two-level three phase PWM inverter, in order to generate a fundamental component, with variable amplitude and frequency of the inverter output voltage, by varying the modulation index and reference frequency. The developed strategy aims to explore the maximum of the available DC bus that supplies the inverter with a minimum level of THD, and uses the minimum number of logic elements LEs available on the FPGA. This technique is verified through simulation results. An experimental power electronics set-up that consists of a DC to AC PWM inverter, a DC supply, and an output LR low pass filter is built and controlled using the Altera DE2 development and education board based on a Cyclone II ALTERA FPGA to test the proposed THIPWM architecture.

Adaptive neural network output feedback control for flexible multi-link robotic manipulators

ABSTRACT In this paper, a novel approach for adaptive control of flexible multi-link robots in the joint space is presented. The approach is valid for a class of highly uncertain systems with arbitrary but bounded dimension. The problem of trajectory tracking is solved through developing a stable inversion for robot dynamics using only joint angles measurement; then a linear dynamic compensator is utilised to stabilise the tracking error for the nominal system. Furthermore, a high gain observer is designed to provide an estimate for error dynamics. A linear in parameter neural network based adaptive signal is used to approximate and eliminate the effect of uncertainties due to link flexibilities and vibration modes on tracking performance, where the adaptation rule for the neural network weights is derived based on Lyapunov function. The stability and the ultimate boundedness of the error signals and closed-loop system is demonstrated through the Lyapunov stability theory. Computer simulations of the proposed robust controller are carried to validate on a two-link flexible planar manipulator.