Three workshops are proposed on Monday 28 October 2013

The fees for each workshop are The fees include attendance and coffee break. (The registration deadline is: October 12, 2013)

Please complete the Registration form and send it to Dr. Mohamed MENAA

For more information, feel free to contact Dr. Mohamed MENAA


Generalized Dynamic Inversion: An Evolving Paradigm for Control System Design

by Prof. Abdulrahman Bajodah, Aeronautical Engineering Department, King Abdulaziz University, Jeddah, Saudi Arabia


The development of Nonlinear Dynamic Inversion (NDI) has contributed substantially towards the advancements in control systems design during the past three decades. The methodology is nowaday simplemented within numerous control systems of the latest state- of the art industrial applications, e.g., civil and military aircraft,spacecraft, electric motors and devices, robots, and chemical plants. Despite of its popularity, NDI has its shortcomings, which are mostly related to the square inversion involved in the methodology. These include singularities of the inverted matrices, cancelations of useful nonlinearities, large control efforts, limited choices of controlled variables due to the zero dynamics, in addition to robustness concerns. The aim of this workshop is to present the newly developed generalized dynamic inversion (GDI) control paradigm, which intends to overcome the drawbacks of classical NDI. In particular, the inversion process that utilizes non-square generalized (pseudo-) inversion will be highlighted, in addition to the associated nullspace parametrization of non-square system matrices. The algebraic and the geometrical features of GDI that facilitate control system design will be emphasized, and the opportunities to employ several well-known control design methodologies in the framework of GDI will be discussed. Finally, the advantages of GDI over classical NDI and other control methodologies will be illustrated by several aerospace,electrical, and mechanical engineering examples.


  1. Overview of classical Nonlinear Dynamic Inversion (NDI)
    • Summary of the NDImethodology and feedback linearization
    • Shortcomings and drawbacks of NDI
  2. Generalized Dynamic Inversion (GDI)
    • Mathematical preliminaries
    • Feedback linearization of non-square systems
    • Nullspace parametrization of nonsquare system matrices
    • Semi-definite Lyapunov functions and Null-control vector design
  3. Augmenting control methodologies in the GDI control design framework
  4. Examples, applications, and open problems

Bond Graph for Integrated Design of Engineering Systems



Bond graph methodology is a graphical language unified for all physical domains. It is a performing tool well adapted for innovating product designing or improvement of existing ones. This is why this multi-physic modelling and simulation tool is widely used by industrials not only for modelling but also for integrated design of intelligent systems. In the present workshop presented by industrial and academic experts, it will be shown how the bond graph tool can be used for dynamic and automated modelling, health monitoring and design of supervision design of test rigs applied to aeronautic systems. The presentations will be illustrated by industrial applications and dedicated software.


  1. 9h-12h30h. Bond Graphs for Modeling and Control Analysis: Application to Mechatronic Systems.
    Professor G. Dauphin Tanguy, Ecole Centrale de Lille, France.

    The basic principles and tools of the bond graph methodology will be presented. It will be pointed out how causality properties are shown up on the bond graph by means of the causal stroke, allowing the building of mathematical models from the bond graph and the implementation of graphical procedures for structural analysis purposes. Several industrial applications will be briefly shown up.

  2. 14h-17h. Bond Graphs for Design of Supervision Systems: Application to Process Engineering.
    Professor B. Ould Bouamama, Ecole Polytechnique de Lille, France.

    The term Supervision means a set of tools and methods used to operate an industrial process in normal and faulty situation. It will be shown how the bond graph tool because of its causal and structural properties can be used for modeling and diagnosis algorithms generation. A specific software developed by the author will be presented and tested to a pedagogical and real processes to be monitored.

  3. 17h.. Discussion

Industrial Parallel Robotics: New Challenges in Design and Control


Nowadays, all around the world, robots are used to perform a very large variety of manipulation tasks in a vast range of industries. They are used in assembling, cutting, spraying, painting, welding, palletizing, polishing, packaging, and more applications. Despite the existence of different types of robots, parallel structures are increasingly used in industry. They are commonly well known for their stiffness because of the entirely closed kinematic chains and their lightness giving them the ability to be very fast. This workshop aims to present Parallel Robotics as a real opportunity for industry through the presentation of different kinds of existing and possible applications. The topics emerging from recent advances in optimization, design and control of this kind of structures are particularly pointed out to emphasize on the existing results and the new challenges to be addressed. Optimization and dimensional synthesis of robots are capital for the design of an industrial structure. Different procedures and methods exist. A synthesis methodology in which the problem is expressed in terms of multi-objective optimization is pointed out during this workshop. Several criteria of performance such as workspace optimization, kinetostatic performances, stiffness, repeatability and dynamic dexterity are taken into account. Interesting results through different use cases will be presented and discussed. Special attention will be given to parallel manipulators with actuation redundancy, which provide very homogenous performances all over their workspace. When dealing with control of parallel robots, interesting issues include fast pick-and-place applications. Indeed, looking for short-cycle times in robotized manipulations means obviously to look for short motion and stabilization times, while guaranteeing the robustness of performances with respect to disturbances and changes/uncertainties in operational conditions. The topic of parallel robot control is particularly detailed. Different control strategies and particularly, the design of an adaptive controller for extremely fast pick-and-place tasks (up to 100G of acceleration) are pointed out. Experimental results, for different case studies, will be presented and discussed.


09h00-11h00: Parallel robotics: opportunities for industry through different case studies. By Dr. Mohamed Bouri, LSRO, EPFL, Lausane, Swisse
11h00-12h00: Parallel robotics in TECNALIA: From design to industrial prototype. By Dr. Lotfi Chikh, TECNALIA, France
14h00-15h00: Control of parallel robots for high accelerations: challenges and solutions. By Dr. Ahmed Chemori, LIRMM, CNRS/University Montpellier 2, France
ARROW - A Redundantly Actuated Parallel Manipulator, Capable of High Acceleration and High Repeatability. By Dr. Sébastien Krut, LIRMM, CNRS/Université Montpellier 2, France
16h00-17h00: Multiobjective optimization for design of parallel robots. By Dr. Ridha Kelaiaia, University of Skikda, Algeria





Technical Co-Sponsors

Friends of the Conference