5R16. Modeling, Identification and Control of Robots. - W Khalil (Ecole Centrale de Nantes, France) and E Dombre (Robotics Dept LIRMM, UMR CNRS, France). Hermes Sci Publ, Paris. Distributed in USA by Taylor & Francis Publ, New York NY. 2002. 480 pp. ISBN 1-56032-983-1. $149.00.

Reviewed by ML Nagurka (Dept of Mech and Indust Eng, Marquette Univ, PO Box 1881, Milwaukee WI 53201-1881).

This book is a revised and augmented edition of the French version, Mode´lisation, Identification et Commande Des Robots, published by Herme´s in 1999, whose first edition was published in 1988. The authors consider this book to be the third edition, as it has been substantially modified and updated. The book contains 15 chapters, 11 appendices, and an extensive list of references. It unfolds as follows.

Chapter 1 is an introduction to the terminology and presents general information and definitions of core concepts, such as kinematic chains, types of joints, configuration versus task space, redundancy, singular configurations, architectures of robot manipulators, and robot characteristics.

Chapter 2, Transformation matrix between vectors, frames and screws, sets out the basic mathematical tools used in robot modeling, including homogeneous and differential transformations as well as screws, twists, and wrenches.

Geometric representations of simple open chain robots are developed in Chapter 3, Direct geometric model of serial robots. A variation of the Denavit-Hartenberg notation, called the Khalil-Kleinfinger notation, is introduced to describe robot geometry. The authors claim that their notation also handles the description of complex chains with tree structures or closed loops.

In Chapter 4, Inverse geometric model of serial robots, three approaches are presented: the Paul method, which can be used for most industrial robots; the Pieper method, which deals with six degree-of-freedom robots having three prismatic joints or a spherical joint; and the Raghavan-Roth method, which is suitable for six degree-of-freedom robots with general geometry.

After developing efficient methods for calculating the Jacobian matrix, Chapter 5, Direct kinematic model of serial robots, presents several analysis-oriented issues, including robot workspace, determination of the degrees-of-freedom, velocity and force ellipsoids, and twist-wrench duality. The kinematic model can also be used to find a numerical solution to the inverse geometric problem. This is the topic of Chapter 6, Inverse kinematic model of serial robots, where solution techniques are provided for regular, singular, and redundant robot configurations.

Chapter 7, Geometric and kinematic models of complex chain robots, examines models of complex chain robots with tree and closed chain structures. The problem of solving the constraint equations of closed loop robots is treated using geometric constraint equations and kinematic constraint equations.

Parallel structured robots are the subjects of Chapter 8, Introduction to geometric and kinematic modeling of parallel robots, where their architectures and features are presented.

Chapters 9 and 10 tackle issues of dynamic modeling. Simple open chains are considered in Chapter 9, Dynamic modeling of serial robots, and complex kinematic chains in Chapter 10, Dynamics of robots with complex structure. Both Lagrangian and Newton-Euler formulations are developed to obtain the robot equations of motion. The determination of the minimum inertial parameters, also referred to as base inertial parameters, is carried out using a direct symbolic method and by a numerical method, based on a QR decomposition. The number of operations of the inverse dynamic model is minimized by using the base parameters and customized symbolic programing techniques. The chapters discuss on-line implementation issues and give different methods for the direct dynamic model computation, including a method that avoids inverting the inertia matrix.

Chapters 11 and 12 focus on identification of geometric and dynamic parameters, respectively. In Chapter 11, Geometric calibration of robots, various calibration methods are offered, including those that require information from external sensors and those that are autonomous. A short subsection introduces the active field of research into parallel robot calibration. In Chapter 12, Identification of the dynamic parameters, several methods (all linear in the dynamic parameters) for identification of dynamic models and energy models are introduced.

Chapter 13 covers Trajectory generation. Beginning with point-to-point trajectories in the joint space and in the task space, the chapter, then examines the problem of adding intermediate points. The topic of trajectory generation on a continuous path is also treated briefly.

Robot control issues are addressed in the last two chapters. Chapter 14, Motion control, covers PID control, computed torque control, passive control and adaptive control, whereas Chapter 15, Compliant motion control, explores passive control, impedance control, hybrid force-position control, and hybrid external control.

The chapters are followed by 11 appendices (over 50 pages) that provide details of mathematical methods and examples of computations (eg, solutions of inverse robot equations, computations of parameters, control laws, stability analyses), a bibliography of more than 400 references, and ends with an index.

It is difficult to compete in the field of robot texts and reference books. Several books cover comparable topics, including the influential Introduction to Robotics: Mechanics and Control, by J Craig (2nd edition, Addison-Wesley, 1989), the extensive and more applied Introduction to Robotics, by P McKerrow (Addison-Wesley, 1991), as well as the more current Robot Analysis: The Mechanics of Serial and Parallel Manipulators, by L-W Tsai (Wiley, 1999). In this reviewer’s opinion, the book Modeling, Identification and Control of Robots is a welcome addition to these books.

The book is primarily a mathematical treatise that unfolds logically and covers a wide range of accepted topics in robotics. It is less of a reference for those seeking information about robotic applications. Given its analytical rigor it may be beyond a technician-level book and more suitable for one with sufficient mathematical skills and savvy. The book contains a wealth of information and would be appropriate as an upper-level undergraduate or graduate text for engineering courses.

In closing, this is a comprehensive, well-written, and pleasing book that contains a broad range of material related to robot modeling, identification, and control. The layout is logical, the writing style smooth, and the figures, although more might be warranted, are clear, black-and-white, schematic-type drawings. The book could be strengthened by the inclusion of more example problems, as well as discussion and material of a more applied nature. Its attention to implementation issues, such as the computational burden in carrying out robotic related calculations, is commendable. It has the rigor and completeness to make it appropriate as a textbook for an advanced engineering course, as well as for anyone seeking information in the field. The book is clearly a contribution and is recommended.