Publications

Some of my recent publications, followed by their abstracts and links to the postscript files.
On Goursat Normal Forms, Prolongations, and Control Systems
D. Tilbury and S. Sastry

The Multi-Steering Trailer System: A Case Study in Goursat Normal Forms and Prolongations.
D. Tilbury and S. Sastry

A Multi-Steering Trailer System: Conversion into Chained Form using Dynamic Feedback
D. Tilbury, O. Sordalen, L. Bushnell, and S. Sastry

Extended Goursat Normal Forms with Applications to Nonholonomic Motion Planning
L. Bushnell, D. Tilbury, and S. Sastry

Trajectory Generation for the N-Trailer Problem using Goursat Normal Form
D. Tilbury, R. Murray, S. Sastry

Steering Three-Input Chained Form Nonholonomic Systems: The Fire Truck Example
L. Bushnell, D. Tilbury and S. Sastry

Steering a Three-Input Nonholonomic System Using Multirate Controls
D. Tilbury and A. Chelouah

Stabilization of Trajectories for systems with nonholonomic contraints
G. Walsh, D. Tilbury, S. Sastry, R. Murray, and J-P. Laumond


On Goursat Normal Forms, Prolongations, and Control Systems

D. Tilbury and S. Sastry

UC Berkeley ERL Memo Number M94/16.
To appear at IEEE Conference on Decision and Control, December 1994.
In this paper, we present the method of exterior differential systems for analyzing nonlinear systems. We present conditions for converting Pfaffian systems into Goursat normal form, and for converting control systems into Brunovsky form. All of the existing results on feedback linearization for control systems can be restated in the language of Pfaffian systems. Our conditions for linearizing control systems using dynamic extension are closer to necessity than those which currently exist in the literature.

ERL Memo (28 pages postscript)
Conference version (6 pages postscript)


The Multi-Steering Trailer System: A Case Study in Goursat Normal Forms and Prolongations.

D. Tilbury and S. Sastry

To appear in International Journal of Robust and Nonlinear Control, 1995.
In this paper, wheeled mobile robots towing trailers are used as examples of systems which can be converted into extended Goursat normal form. For some configurations of this class of systems, a prolongation of the Pfaffian system is necessary before the Goursat form can be achieved. The advantage of the Goursat form is that once the kinematic constraints are in this normal form, the point-to-point path planning problem can be solved in a straightforward way.

(23 pages postscript)


A Multi-Steering Trailer System: Conversion into Chained Form using Dynamic Feedback

D. Tilbury, O. Sordalen, L. Bushnell, and S. Sastry

UC Berkeley ERL Memo Number M93/55.
Preprints of the IFAC Symposium on Robot Control, September 1994.
To appear in IEEE Transactions on Robotics and Automation, 1995.
In this paper, we examine in detail the kinematic model of an autonomous mobile robot system consisting of a chain of steerable cars and passive trailers, connected together with rigid bars. We define the state space and kinematic equations of the system, modeling the pair of wheels on each axle as able to roll but not slip. We then investigate how this system of kinematic equations may be converted into a multi-input chained form. The advantages of the chained form are that many methods are available for the open-loop steering of such systems as well as for point-stabilization.

In order to convert the system to this multi-input chained form, we use dynamic state feedback. We draw some motivation from the very simple example of a kinematic unicycle and the relationships of the angular velocities therein, and we show how the dynamic state feedback that we use corresponds to adding, in front of the steerable cars, a chain of virtual axles which diverges from the original chain of trailers.

We briefly discuss how some of the methods which have been proposed for steering and stabilizing two-input chained form systems can be generalized to multi-chained systems. For concreteness, we also present two different example systems: a fire truck (three axles) and a five-axle, two-steering system. Simulation results for a parallel-parking maneuver for the five-axle system are included in the form of margin movies.

(33 pages postscript)
Short conference version: (6 pages postscript)


Extended Goursat Normal Forms with Applications to Nonholonomic Motion Planning

L. Bushnell, D. Tilbury, and S. Sastry

IEEE Conference on Decision and Control, 1993.
The theme of this paper is the generalization of Goursat normal forms for Pfaffian systems with co-dimension greater than two. There are necessary and sufficient conditions for the existence of coordinates that transform a Pfaffian system with co-dimension greater than two into an extended Goursat normal form, which is the dual of the multiple-chain, single-generator chained form mentioned in our earlier work. In this paper, we concentrate on how to find such coordinate transformations for multi-steering, multi-trailer mobile robot systems so that we can use available steering and stabilization algorithms for nonholonomic motion planning. We present a methodology for constructing a coordinate transformation and apply it to the example of a five-axle, two-steering mobile robot.

(12 pages postscript)


Trajectory Generation for the N-Trailer Problem using Goursat Normal Form

D. Tilbury, R. Murray, S. Sastry

UC Berkeley ERL Memo Number M93/12.
IEEE Conference on Decision and Control, 1993.
To appear in IEEE Transactions on Automatic Control, 1995.
In this paper, we develop the machinery of exterior differential forms, more particularly the Goursat normal form for a Pfaffian system, for solving nonholonomic motion planning problems, ie planning problems with non-integrable velocity constraints. We apply this technique to solving the problem of steering a mobile robot with n trailers. We present an algorithm for finding a family of transformations which will display the given system of rolling constraints on the wheels of the robot with n trailers in the Goursat canonical form. Two of these transformations are studied in detail. The Goursat normal form for exterior differential systems is dual to the so-called chained form for vector fields that we have studied in our earlier work. Consequently, we are able to give the state feedback law and change of coordinates to convert the N-trailer system into chained form. Three methods for steering chained form systems using sinusoids, piecewise constants and polynomials as inputs are presented.

The motion planning strategy is therefore to first convert the N-trailer system into chained form, steer the corresponding chained form system, then transform the resulting trajectory back into the original coordinates. Simulations and frames of movie animations of the N-trailer system for parallel parking and backing into a loading dock using this strategy are also included.

(52 pages postscript)
Short conference version: (7 pages postscript)


Steering Three-Input Chained Form Nonholonomic Systems: The Fire Truck Example

L. G. Bushnell, D. M. Tilbury and S. S. Sastry

1993 European Control Conference.
To appear in the International Journal of Robotics Research, 1995.
In this paper, we steer nonholonomic systems with linear velocity constraints represented mathematically in a special form, called chained form. We observe that chained form systems can be steered from an initial configuration to a final configuration with sinusoidal inputs. The controller we use is open loop and no special provisions are made for obstacle avoidance. Sufficient conditions are presented for converting a three-input system with nonholonomic velocity constraints into a ``two-chain, single-generator chainedform.'' An algorithm is stated that constructs the sinusoidal control inputs to steer this system from any initial configuration to any desired final point. Our example of a three-input nonholonomic system is a firetruck, or tiller truck. In this three-axle system, the control inputs are the steering velocities of both the front and rear wheels of the truck and the driving velocity of the truck. Simulation results are given for the familiar parallelparking problem and other trajectories.In tIn this paper we examine a multi-rate control scheme for nonholonomic path planning using constant control inputs over different time periods. For chained systems, an exact point-to-point trajectory is generated. Simulation results are presented for a three-input system, and comparisons are made with a sinusoidal method for path planning.

(24 pages postscript)
Short Conference version: (6 pages postscript)


Steering a Three-Input Nonholonomic System Using Multirate Controls

D. Tilbury and A. Chelouah

1993 European Control Conference.
In this paper we examine a multi-rate control scheme for nonholonomic path planning using constant control inputs over different time periods. For chained systems, an exact point-to-point trajectory is generated. Simulation results are presented for a three-input system, and comparisons are made with a sinusoidal method for path planning.

(4 pages postscript)


Stabilization of Trajectories for Systems with Nonholonomic Constraints

G. Walsh, D. Tilbury, S. Sastry, R. Murray, and J-P. Laumond

IEEE Transactions on Automatic Control, January, 1993.
IEEE International Conference on Robotics and Automation, 1991.
A new technique for stabilizing nonholonomic systems to trajectories is presented. It is well known that such systems cannot be stabilized to a point using smooth static state feedback. In this paper we suggest the use of control laws for stabilizing a system about a trajectory, instead of a point. Given a nonlinear system and a desired (nominal) feasible trajectory, the paper gives an explicit control law which will locally exponentially stabilize the system to the desired trajectory. The theory is applied to several examples, including a car-like robot.

(17 pages postscript)


Dawn Tilbury (dawnt@eecs.berkeley.edu) / November 1 , 1994