Arab Ali-Chérif

A robust adaptive control of a parallel robot

The work presented in this article deals with the robust adaptive control tracking of a 6 degree of freedom parallel robot, called C5 parallel robot. The proposed approach is based on the coupling of sliding modes and multi-layers perceptron neural networks (MLP-NNs). It does not require the inverse dynamic model for deriving the control law. The MLP-NN is added in the control scheme to estimate the gravitational and frictional forces along with the non-modelled dynamic effects. The nonlinearity problem, present in neural networks, is resolved using Taylor series expansion. The proposed approach allows to adjust the parameters of neural network and sliding mode control terms by taking into account a reference model and the closed-loop stability in the Lyapunov sense. We implemented our approach on the C5 parallel robot of LISSI laboratory and performed experiments to observe its effectiveness and the robust behaviour of the controller against external disturbances.

Multi-robot heuristic goods transportation

In this paper, we consider the issue of transporting a certain number of goods by a team of mobile robots. The target is to minimize the total transportation time and keep a low energy consumption of the intelligent agents on assuring security and quality during the transportation process. The pivotal issue needs to be solved is how to assign tasks to individual robots in a more reasonable and efficient way. We present a novel solution by using an empirical-based heuristic planning strategy for the goods transportation by multiple robots. In contrast to previous approaches, this strategy is designed to plan the transportation task for each individual robot by estimating the production rate of goods based on multi-robot coordination. Our approach has been implemented and evaluated in simulation. The experimental results demonstrate that the completion time of the whole transportation mission can be significantly reduced and the energy consumption of robots can be kept at a low level of our heuristic planning strategy compared with the previous approach.

Smart Wheelchair design and monitoring via wired and wireless networks

Several works for handicapped peoples help and assistance have been investigated in the last decade and particularly those concerning hardware and software architectures design related to automated wheelchairs. Thus, make a wheelchair intelligent and autonomous, leads to develop new strategies taking into account in one hand, environment dynamics, and in other hand, recent communication technologies such as ad hoc radio networks, sensor networks, wireless mesh networks, etc. The aim of this paper is to present new hardware and software architecture of an intelligent system called LIASD-Wheelchair. This system is a kind of home welfare tools used to assists and help the handicapped and elderly people to gain mobility and lead to independent life. The developed wheelchair moves fully autonomously according to environment changes and uses a multimode communication for its maneuverability and localization via Wi-Fi communication network. To test the effectiveness of the developed system, an experiment is designed in this respect.

Combined multi-layer perceptron neural network and sliding mode technique for parallel robots control : An adaptive approach

In this paper, an adaptive control of a parallel robot is proposed for trajectory tracking problems. This approach is based on adaptive multi-layer perceptron (MLP) neural network and sliding mode technique. The aim of this study is to design a robust controller with respect to external disturbances in order to improve the trajectory tracking. In fact, an adaptive MLP neural network is developed to estimate the gravitational force, frictions and other dynamics. To overcome the non-linearity problem presented in the neural network, we used the Taylor series expansion. The control law combining a neural network and sliding mode is synthesized in order to attract states model to the sliding surface. All adaptation laws of neural parameters and sliding mode term are based on the stability of the closed loop system in the Lyapunov sense. This approach has been implemented on a C5 parallel robot, and the experimental results show the effectiveness of the proposed method in presence of external disturbances.

A stable adaptive force/position controller for a c5 parallel robot: A neural network approach

This paper presents an adaptive force/position controller for a parallel robot executing constrained motions. This controller, based on an MLPNN (or Multi-Layer Perceptron Neural Network), does not require the inverse dynamic model of the robot to derive the control law. A neural identification of the dynamic model of the robot is proposed to determine the principal components of the MLPNN input vector. The latter is used to compensate the dynamic effects arising from the robot-environment interaction and its parameters are adjusted according to an adaptation law based on the Lyapunov-analysis methodology. The proposed controller is evaluated experimentally on the C5 parallel robot. This method is capable of tracking accurately the force/position trajectories and its stability robustness is proved.

Hierarchical routing approach-based energy optimization in wireless sensor networks

We propose a critical improvement of the LEACH (Low-Energy Adaptive Clustering Hierarchy) routing protocol for the optimization of the energy consumption as well as memory occupation of Wireless Sensor Network (WSN). Our protocol LEACH-M uses a cascade of clustering algorithms, every step of which chooses the next one. We expect at best an improvement of 5% of energy consumption and since the lifetime of the application is itself improved the amount of data routed is improved in the same amount. Of course, testing our approach using it along a comparative study with the standard LEACH confirms our expectation.

Routing Information for Energy Management in WSNs

The main objective in a WSN application is to guarantee the transmission of information between different sensor nodes according to a pre-established routing protocol. Improving performance in terms of longevity, connectivity and robustness requires some consideration of constraints such as energy consumption, bandwidth and the optimal use of calculation and memory resources. The conception and implementation of a routing protocol can be influenced by several factors that must be addressed before it is possible to ensure efficient data communication. Among these factors, we can refer to the deployment of nodes in operational environments. Being dependent on the kind of applications under consideration, the deployment phase can greatly affect routing performances. It can be determinist and, in this case, sensor nodes are placed manually and the data are transmitted according to pre-determined routes from source nodes to the destination. This can also be done randomly by haphazardly distributing the nodes, as in ad hoc networks. In this case, an optimal classification of the network in clusters is highly necessary in order to guarantee connectivity and enable efficient energy management. Various studies have shown that the communication between nodes in restricted areas, i.e. clusters, provides better results in terms of energy consumption and bandwidth, and information routing can be guaranteed in this case through multiple hops. Moreover, the lifetime of a sensor node is strongly linked to the lifetime of its battery and the malfunctioning of one or several sensor nodes, due to battery failures or a lack of energy resources, can cause significant topological changes that require the network to be completely reorganized in order to ensure the rerouting of data. It is therefore important to consider the issue of how to manage energy consumption without a loss of performance and network precision, particularly in the case of a multi-hop WSN where each node can act as either a data transmitter or a simple router.

Human Ear recognition based on Multi-scale Local Binary Pattern descriptor and KL divergence

This paper presents a novel human ear recognition approach based on Multi-scale Local Binary Pattern (MLBP) descriptor to enhance the recognition performance. The proposed method includes the following two steps: (i) the feature extraction step that computes the MLBP descriptor-based features from human ear images, and (ii) the matching process that uses the Kullback Leibler (KL) distance to capture efficiently the similarities/dissimilarities between the feature vectors and then make a decision. The proposed method is performed using the IIT Delhi Ear database and then compared to the state-of-the-art methods. The results obtained have shown that the proposed method achieves satisfying identification performances up to 95% in terms of rank-1 identification rate.

Algorithms of control by thought in robotics: Active and passive BMIs based on prior knowledge

One of the objectives of the control using the human thought is to make useful robotic systems for persons with high dependency (quadriplegics, paraplegics, etc.). When the human subject is not able to move his limbs, upper or lower, he is no longer able to perform basic and necessary tasks in his daily life. Recently, robotic systems have reached a very advanced level. For example, humanoid robots have become able to walk, recognize and carry objects simultaneously. On the other hand, wearable robots or exoskeletons can help dependent human subject to move and perform tasks previously difficult to imagine. Of course, all these robotic systems cannot perform these tasks except if they are fitted with advanced control schemes. To make these robotic systems, having already some intelligence, more useful, many researchers have studied the problem of controllers based on the user thought. The real challenge is to translate/classify correctly the thought of the user into robotic actions. When the brain activities are not correctly classified or the action thought by the user is not quite performed, it is important to discover it at time. This allows us to update the classifier/controller parameters in order to interpret more precisely the brain activities concerning the following action. This paper deals with looking for relevant prior knowledge that can anticipate any classification error. Thereafter, we propose some reflections regarding the control of robots by passive thought. Our analysis and results are based on the brain machine interface (BMI) using the Steady State Visual Evoked Potentials technique (SSVEP).

Energy Management in Wireless Sensor Networks

Over the last few years, the technological advances in wireless sensor network (WSN) applications have sparked great curiosity and a growing interest among both users and manufacturers, as well as in the research community. Significant challenges have been overcome to ensure their implementation by addressing problems arising from deployment and connectivity, and from routing and securing information, although much remains to be done at the energy management stage. A WSN is made up of a set of sensor nodes, using supply devices or batteries to operate and interconnected via radio links to ensure data reception, processing and transmission. Increasing the autonomy of sensors and extending the network lifetime can therefore be considered among the main objectives by examining interesting methods and studies that optimize energy consumption, and suggesting mechanisms to improve it. These mechanisms can involve several action levels which can range from the deployment stage to the information exploitation stage.