Safe Robot Navigation Among Moving and Steady Obstacles

Safe Robot Navigation Among Moving and Steady Obstacles

First Edition

Alexey S. Matveev

Andrey V. Savkin

Michael Hoy

Chao Wang

Table of Contents

Cover image

Title page

Copyright

Preface

Abbreviations

Frequently used notations

1: Introduction

Abstract

1.1 Collision-free navigation of wheeled robots among moving and steady obstacles

1.2 Overview and organization of the book

1.3 Sliding mode control

1.4 Experimental equipment

2: Fundamentals of sliding mode control

Abstract

2.1 Introduction

2.2 Sliding motion

2.3 Filippov solutions

3: Survey of algorithms for safe navigation of mobile robots in complex environments

Abstract

3.1 Introduction

3.2 Problem considerations

3.3 Model predictive control

3.4 Sensor-based techniques

3.5 Moving obstacles

3.6 Multiple robot navigation

4: Shortest path algorithm for navigation of wheeled mobile robots among steady obstacles

Abstract

4.1 Introduction

4.2 System description and main assumptions

4.3 Off-line shortest path planning

4.4 On-line navigation

4.5 Computer simulations

4.6 Experiments with a real robot

5: Reactive navigation of wheeled robots for border patrolling

Abstract

5.1 Introduction

5.2 Boundary following using a minimum distance sensor: System description and problem statement

5.3 Main assumptions of theoretical analysis

5.4 Navigation for border patrolling based on minimum distance measurements

5.5 Computer simulations of border patrolling with a minimum distance sensor

5.6 Boundary following with a rigidly mounted distance sensor: Problem setup

5.7 Assumptions of theoretical analysis and tuning of the navigation controller

5.8 Boundary following with a rigidly mounted sensor: Convergence of the proposed navigation law

5.9 Computer simulations of border patrolling with a rigidly mounted distance sensor

5.10 Experiments with a real robot

6: Safe navigation to a target in unknown cluttered static environments based on border patrolling algorithms

Abstract

6.1 Navigation for target reaching with obstacle avoidance: Problem statement and navigation strategy

6.2 Assumptions of theoretical analysis and convergence of the navigation strategy

6.3 Computer simulations of navigation with obstacle avoidance

7: Algorithm for reactive navigation of nonholonomic robots in maze-like environments

Abstract

7.1 Introduction

7.2 Problem setup and navigation strategy

7.3 Assumptions of theoretical analysis and tuning the navigation law

7.4 Convergence and performance of the navigation law

7.5 Simulations and experiments with a real wheeled robot

AppendixA Appendix: Proofs of Proposition 4.1 and Lemmas 4.6 and 4.7

8: Biologically-inspired algorithm for safe navigation of a wheeled robot among moving obstacles

Abstract

8.1 Introduction

8.2 Problem description

8.3 Navigation algorithm

8.4 Mathematical analysis of the navigation strategy

8.5 Computer simulations

8.6 Experiments with a laboratorial wheeled robot

8.7 Algorithm implementation with a robotic wheelchair

8.8 Algorithm implementation with a robotic motorized hospital bed

9: Reactive navigation among moving and deforming obstacles: Problems of border patrolling and avoiding collisions

Abstract

9.1 Introduction

9.2 System description and border patrolling problem

9.3 Navigation for border patrolling

9.4 Main assumptions

9.5 Main results concerning border patrolling problem

9.6 Illustrative examples of border patrolling

9.7 Navigation in an environment cluttered with moving obstacles

9.8 Simulations

9.9 Experimental results

10: Seeking a path through the crowd: Robot navigation among unknowingly moving obstacles based on an integrated representation of the environment

Abstract

10.1 Introduction

10.2 Problem description

10.3 Navigation algorithm

10.4 Mathematical analysis of the navigation strategy

10.5 Computer simulations

10.6 Experiments with a real robot

11: A globally converging reactive algorithm for robot navigation in scenes densely cluttered with moving and deforming obstacles

Abstract

11.1 Introduction

11.2 Problem setup

11.3 The navigation algorithm

11.4 Collision avoidance

11.5 Achieving the main navigation objective

11.6 Illustrations of the main results for special scenarios

11.7 Simulations

12: Safe cooperative navigation of multiple wheeled robots in unknown steady environments with obstacles

Abstract

12.1 Introduction

12.2 Problem statement

12.3 Proposed navigation system

12.4 Simulation results

12.5 Experimental results with wheeled robots

Bibliography

Index

Copyright

Butterworth Heinemann is an imprint of Elsevier

The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK

225 Wyman Street, Waltham, MA 02451, USA

© 2016 Elsevier Inc. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices

Knowledge and best practice in this field are constantly changing. As new research and experience broaden our understanding, changes in research methods, professional practices, or medical treatment may become necessary.

Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein. In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

For information on all Butterworth Heinemann publications visit our website at http://store.elsevier.com/

ISBN: 978-0-12-803730-0

Publisher: Joe Hayton

Acquisition Editor: Sonnini Yura

Editorial Project Manager: Mariana Kuhl Leme

Editorial Project Manager Intern: Ana Claudia A. Garcia

Production Project Manager: Sruthi Satheesh

Marketing Manager: Louise Springthorpe

Cover Designer: Victoria Pearson

Preface

Alexey S. Matveev; Andrey V. Savkin; Michael Hoy; Chao Wang

Collision-free navigation of mobile robots in environments cluttered with both static and dynamic obstacles is a fundamental problem of robotics. In order to safely operate in a priori unknown complex environments, autonomous unmanned vehicles should be equipped with an automatic navigation system that allows them to avoid collisions with obstacles. Moreover, the capacity of safe navigation among moving and steady obstacles is a key issue for most-real life applications of mobile robotics. Despite extensive research, this fundamental problem still represents a real challenge. This book is primarily a research monograph that presents, in detail and in a unified manner, recent advancements in this area. In doing so, the book extends its scope to obstacles undergoing general motions, including rotations and deformations, i.e., changes of the shape and size, and gives considerable attention to reactive algorithms and rigorous mathematical studies of the proposed navigation solutions, showing the interplay between mathematics and robotics. At the same time, the theoretical results are systematically confirmed via computer simulation studies and real-world experimental validation of the developed algorithms with various mobile robots.

The main intended audience for this monograph encompasses students at postgraduate and graduate levels, as well as professional researchers and industry practitioners working in a variety of areas, notably robotics, control engineering, computer sciences, and applied mathematics, who have an interest in the fascinating and cutting-edge discipline of mobile robotics.

This book is essentially self-contained; the only prerequisite is familiarity with basic undergraduate level mathematics. The theories and algorithms presented are discussed fully and illustrated by examples; the proposed navigation solutions are computationally efficient and can be easily instrumented in practice. Therefore, some basic experience in robotics and programming is enough to implement these algorithms directly from the book. We hope readers find this volume useful, clear, and interesting, gain a deeper insight into the challenging field of collision-free robot navigation, and are encouraged to explore it further by following a similar road.

The material in this book is the result of the joint research of the authors from 2009 to 2015. Some of its parts have separately appeared in journal and conference papers. The manuscript integrates them into a unified whole, highlights their coherence and connections among them, supplements them with new original findings of the authors, and presents the entire material in a systematic way.

In preparation of this work, the authors wish to acknowledge the financial support they received from the Australian Research Council; Alexey Matveev gratefully acknowledges the support from RSF 14-21-00041 in preparation of the book except for Chapter 7, and the support by grant 6.38.230.2015 from the Saint Petersburg State University in preparation of Chapter 7. The authors are also grateful for the support they received throughout the production of this book from the School of Electrical Engineering and Telecommunications at the University of New South Wales, Sydney, Australia and the Faculty of Mathematics and Mechanics at the Saint Petersburg State University, Saint Petersburg, Russia.

Furthermore, Alexey Matveev is grateful for the enormous support he has received from his wife Elena and daughter Julia. Andrey Savkin is indebted to the endless love and support he has received from his wife Natalia, son Mikhail, and daughter Katerina. Michael Hoy is thankful for the continuous support of his loving fiancée Reneé. Chao Wang is thankful for the support he has received from his family and friends.

Abbreviations

AM   avoidance maneuver

APF   artificial potential field

DMPC   decentralized model predictive control

DRCA   distributed reactive collision avoidance

GPT   generator of planned trajectories

GPPT   generator of presumable planned trajectories

GRPPT   generator of refined presumable planned trajectories

IT   initial turn

LiDAR   light detection and ranging

LQG   linear quadratic Gaussian

LTM   longitudinal tracking module

MPC   model predictive control

NLVO   nonlinear velocity obstacle

PD   proportional derivative

PFM   potential field method

PGL   proposed guidance law

PTM   path-tracking module

RCA   reciprocal collision avoidance

RHC   receding horizon control

RPI   robustly positively invariant

SMEC   sliding motion along the equidistant curve

SMT   straight motion to the target

TPM   trajectory planning module

TTM   trajectory tracking module

UAV   unmanned aerial vehicle

VO   velocity obstacle

VOA   velocity obstacle approach

VSS   variable structure system

i.e.   that is (Latin: id est)

e.g.   for example (Latin: exempli gratia)

etc.   and so forth, and so on, continuing in the same way (Latin: etcetera)

Frequently used notations

:=  

is equal by definition, is set to be

⇒  

entailment: "A⇒B" means A implies B

⇔  

equivalence: "A⇔B" means A equivalent to B

≡  

identity: "A ≡ B" means A and B are identical, i.e., are the same

∧  

logical and

∨  

logical or

∀  

for all

∃  

exists, there exists

  the relation is true by virtue of A; here = can be replaced by ≤,⇒ and any other relation

{a1,..., an}   the set composed by the elements a1,..., an

∅  

the empty set

{a : C(a)}   the set of all a for which the claim C(a) is true; e.g., {a : a ≤ 1} is the set of all a not exceeding 1

A ∪ B   the union of the sets A and B

A ∩ B   the intersection of the sets A and B

A \ B   the difference of the sets A and B: the set of all points from A that do not belong to B

f (·)   dot in parentheses signals that the preceding symbol denotes a function

  the field of real numbers

T  

transposition

col{a1,..., an}   the column composed of the listed elements: col{a1,..., an} = (a1,..., an)T

n   the Euclidean space of all column-vectors a = col{a1,..., an} with real entries

  the standard inner product in a Euclidean space:

  the standard inner product in a Euclidean space:

||·||  

max{a1,..., an}   the maximum among the listed numbers

min{a1,..., an}   the minimum among the listed numbers

sgn  

the signum function: sgn a = 1 if a > 0, sgna = −1 if a ‹ 0, and sgn 0 = 0

[a]+   the positive portion of a real number a: [a]+ := a if a ≥ 0, otherwise [a]+ := 0

[a]-   the negative portion of a real number a: [a]- := a if a ≥ 0, otherwise [a]- := 0

a   the integer ceiling of a real number a: the least integer that is no less than a

a   the integer floor of a real number a: the largest integer that is no greater than a

a = b mod c   this means that a − b is divisible by c

ai ↑   the sequence is nondecreasing: ai+1 ≥ ai ∀i;

ai ↑ ∞   the sequence is nondecreasing and grows without limits

t ≈ τ   "t is sufficiently close to τ"

t ≈ ∞   "t is sufficiently large"

t → τ−   "t converges to τ from the left"

t → τ+   "t converges to τ from the right"

[a, b] ⊂ n   the closed segment with end-points a and b: [a, b] = {x : x = (1 − θ)a + θb for some 0 ≤ θ ≤ 1}

[a, b] ⊂ n   the open segment with end-points a and b: (a, b) = {x : x = (1 − θ)a + θb for some 0 ≤ θ ≤ 1}

[a, b] ⊂ n   the semi-open segment with end-points a and b: (a, b] = {x : x = (1 − θ)a + θb for some 0 ≤ θ ≤ 1}

dist [r;D]   the distance from the point r to the set D: dist [r;D] := infr*∈D ||r*||

dist [D′,D″]   the distance between two sets D′,D″: dist [D′,D″]:= infr′∈D′,r″∈D′ ||r′ −r″||

∂D   the boundary of the set D

Ck-smooth   k times continuously differentiable

smooth  

Ck-smooth with large enough k

1

Introduction

Abstract

This introductory chapter broadly describes the scope and purpose of the book, reports on its organization, offers a short introduction to sliding mode approach to the problems covered in the book, and provides an overview of the mobile robots used for experimental verification of the algorithms developed in the text.

Keywords

Robot navigation

Obstacle avoidance

Cluttered environments

Wheeled mobile robots

Collision avoidance

1.1 Collision-free navigation of wheeled robots among moving and steady obstacles

Autonomous navigation of unmanned vehicles is a classic research area in robotics, which gives rise to a whole variety of approaches well documented in literature. Both single and multiple coordinated mobile robots offer great perspectives in many applications, such as industrial, office, and agricultural automation; search and rescue; and surveillance and inspection (we refer the reader to [1, 2] for a more comprehensive list of potential applications). For example, because of lightweights, inexpensive components, and low power consumptions, unmanned aerial and ground vehicles have been used greatly for plenary surveillance and for a variety of applications in hazardous and complex environments to mitigate the risk for humans; see, e.g., [3–6] and references therein. Such applications often involve limitations on communications that require the robotic vehicle to operate autonomously for extended periods of time and distances. In such situations, unmanned vehicles should be equipped with an automatic navigation system by which they can move autonomously and safely operate in populated environments. In all these cases, navigation involves a series of common problems, with collision avoidance in some form being almost universally required.

Moreover, the capacity of safely operating in dynamic and a priori unknown environments is a key issue for most real-life applications of mobile robotics. Despite extensive research, this fundamental problem still represents a real challenge, mostly because of the numerous uncertainties inherent in typical scenarios. This challenge can be much enhanced by deficiency in perception abilities and computational power of the robot, as well as by restrictions on its mobility due to nonholonomic kinematic constraints, limited control range, and under-actuation.

In order to operate in a cluttered environment, an autonomous unmanned vehicle should be able to detect and avoid the enroute obstacles. Typical objectives include, but are not limited to, overall movement in a given direction or reaching a target point through the obstacle-free part of the environment [7]. This may involve bypassing an obstacle, especially a long one, in close range within a safety margin [8]. This maneuver is similar to border patrolling, which mission is of self-interest for many applications. Substantial effort has been made in robotics research for solution of these problems. The rich variety of available relevant algorithms will be specifically surveyed in a special Chapter 3. With a focus on the planning horizon, they can be very broadly classified into global and local planners [9].

Global sensor-based planners use both a priori and sensory information to build a comprehensive model of the environment and then try to find a completed and best possible navigation solution on the basis of this model [7, 10–16]. In effect, this means the environment is assumed to be known to a substantial extent prior to the solution of a navigation task. Within this framework, a variety of techniques has been developed. Their general survey is postponed until Chapter 3 but can also be found in, e.g., [17, 18]; some samples intended to handle dynamic scenes are given by velocity obstacles [17, 19], nonholonomic planners [20], and state-time space [21–23] approaches. Global planners can often be accompanied with guarantees of not only collision avoidance but also achieving a global navigation objective provided that certain general assumptions about the scene are satisfied. On the negative side, global planners are, by and large, computationally expensive and hardly suit real-time implementation. NP-hardness, the mathematical seal for intractability, was established for even the simplest problems of dynamic motion planning [24]. A partial remedy was offered in the form of randomized architectures [25, 26]. At the same time, all global planners are hardly troubled, up to failure in path generation, by unpredictability of the scene, as well as by data incompleteness and erroneousness typical for onboard perception. These disadvantages are shared by hybrid approaches that use global planning as a backbone of navigation [15, 20, 27–30].

Conversely, local planners use onboard sensory data about only a nearest fraction of an unknown environment for iterative re-computation of a short-horizon path [9, 31–33]. This reduces the computational burden toward implementability in real time but makes the ultimate result of iterations an open issue. A detailed overview of the respective obstacle avoidance techniques will be given in Chapter 3. Many of them, e.g., the dynamic window [34, 35], curvature velocity [36], lane curvature [37], boundary following [32, 33], and tangent graph based [16] approaches, treat the obstacles as static. This is a particular case of predictably moving obstacles, which are assumed by methods like velocity obstacles [17, 19], collision cones [38], or inevitable collision states [39, 40]. These methods tend to be computationally expensive, are based on access to the obstacles’ full velocities, and assume a modest or minor rate of their change. However, velocity estimation remains a challenging task in practice, and predictability of the scene ranges from full to none in the real world [41]. A medium level of predictability is that with uncertainty, where non-conservative estimates of future obstacle positions can be put in place of exact prognosis [17, 42]. However, these and other approaches [19, 38–40] take in effect excessive precautions against collisions with obstacles. As a result, they may be stuck in cluttered scenes and tend toward bypassing dense clusters of obstacles as a whole, even if a better and sometimes the only option is a permeating route. In hardly predictable complex environments, safety typically concerns only a nearest future, and its propagation until the end of the experiment is not guaranteed [42]. Some local planners, such as Virtual Force Field [43], Potential Field [44, 45], Vector Field Histogram [46], Certainty Grid [47], Nearness Diagram [48] methods, use elements of global modeling by assuming awareness about the scene above the level given by the current sensory data. A common problem with local methods is that many of them are heuristic and not based on mathematical models such as kinematic equations of the vehicles and nonholonomic constraints on their motion, which is a severe limitation in practice. For dynamic environments, fully actuated robots were mostly studied up to now.

Because of inevitable failure scenarios, a common deficiency of the previous research on local planners is the lack of global convergence results that guarantee achieving the primary objective in dynamic environments [49]. At best, rigorous analysis examined an isolated bypass of an obstacle during which the other obstacles were neglected until the bypass end, with an idea that thereafter, the robot focuses on the main goal. However, in cluttered dynamic scenes, bypasses may be systematically intervened by companion obstacles so that no bypass is completed, whereas the robot almost constantly performs obstacle avoidance. Ultimate goal was left, by and large, beyond the scope of theoretical analysis, especially for cluttered unpredictable environments, like a dense crowd of people. However, it is in these cases that rigorous quantitative delineation between failure and success scenarios is highly important since by its own right, any experimentation is not convincing enough due to horizonless diversity of feasible scenarios. Another deficiency is that moving obstacles were viewed as rigid bodies undergoing only translational motions and often of the simplest shapes (e.g., discs [45, 50–52] or polygons [53, 54]). Finally, assumed awareness of the obstacles often meant access to their possibly invisible parts (in order to determine, e.g., the disc center [45, 50, 51] or angularly most distant polygon vertex [54]) or full velocity [45, 51, 54].

The basic strong and weak points of local planners attain apotheosis at reactive controllers. For them, the planning horizon collapses into a point so that the controller directly converts the current observation into the current control. Purely reactive approaches are exemplified by [51, 55–59], as well as by biologically inspired methods [32, 33, 52, 60, 61]. Examples concerning nonholonomic robots include artificial potential approach, combined with sliding mode control for gradient climbing [50, 53], and kinematic control based on polar coordinates and Lyapunov-like analysis [51]. Some local planners, such as Virtual Force Field [43], Potential Field [44, 45], Vector Field Histogram [46], Certainty Grid [47], Nearness Diagram [62] methods, in fact combine reactive control with elements of global modeling by assuming awareness about the scene above the level given by the snapshot of the sensory data. Another example of such local path planners is the biologically inspired Bugs family of algorithms [11, 12, 63, 64] motivated by bugs crawling along a wall.

This monograph is aimed at overcoming some of the afore-mentioned deficiencies in the previous research on navigation of mobile robots among obstacles. It deals with both steady and moving obstacles; the scope of the book extends to not only arbitrarily shaped obstacles but also obstacles undergoing arbitrary motions, including translations, rotations, and deformations, i.e., changes of sizes and shapes. Except for Chapter 4, where elements of the global planning approach are employed to handle global optimization issues, this research text develops local techniques, with a main focus on reactive algorithms. Unlike many other publications in robotics, the proposed navigation solutions are based on rigorous mathematical theory, and systematic theoretical justification of their convergence and performance can be viewed as a novel and distinguishing feature of this book. Mostly, its chapters deliver a consistent message that even purely reactive navigation laws have a capacity of being supplied with mathematically rigorous guarantees of achieving a global navigation objective. Though we sometimes refer the reader to textbooks for technical mathematical facts, important theoretical developments are given with care.

The theoretical part of the work presented in this monograph is confined to models conventionally used to describe wheeled robots. However, the results of this part extend in effect on any controlled system obeying the same governing equations; some typical but not exhaustive examples include, under certain circumstances, fixed wing aircrafts, torpedo-like underwater vehicles, and spacecrafts. Principal solutions are discussed with respect to these basic models, which imparts the feature of hardware independency to the major messages from this book. At the same time, experimental validation of the proposed navigation solutions is a crucial part in robotics research. In this book, the proposed navigation algorithms were tested with real wheeled unmanned ground vehicles such as laboratorial wheeled robots Pioneer P3-DX, an intelligent autonomous wheelchair, and an autonomous wheeled mobile hospital bed. A detailed description of the employed experimental equipment will be offered in Section 1.4.

Another essential trait of this monograph is systematic account for motion constraints inherent in real-world wheeled robots. Their mathematical models employed in most of the technical chapters explicitly incorporate nonholonomic kinematics, limited control range, and under-actuation; which substantially intensifies the challenge of safe navigation among obstacles. The defiance from this triplet of oppressive properties is somewhat reduced only in three of nine chapters: Chapters 8 and 12 deal with fully actuated robots instead of under-actuated ones, whereas only limitations on the control range are taken into account in Chapter 11.

The organization of this book is problem-oriented, not technique-oriented. Each chapter is relatively self-contained and is devoted to detailed discussion of a certain problem that arises in the rapidly developing area of mobile robotics. On the other hand common ideas and methods thread through chapters so that the entire book can be viewed as an advocate for their usefulness as an effective apparatus of wide applicability. In particular, the monograph is featured by systematic use of switched controllers so that in closed loop, the examined systems fall into the class of hybrid dynamical systems [65–69].

The goal of this monograph is to report on some recent progress in the area of navigation of wheeled robots among steady and moving obstacles, and to present, in an unified form, a computationally efficient, reliable, and easily implementable algorithmic framework for control of mobile robots, so that the ultimate goal of their applications can be achieved with a high degree of certainty. Such a framework is very important in the face of ever-increasing real-world applications of wheeled robots since it carries a potential to transcend technological evolution in mobile robotics.

1.2 Overview and organization of the book

In Chapter 2, we briefly recall a few basic concepts and facts of the classic sliding mode control theory, focusing attention on those that are used in this book.

Chapter 3 provides a review of techniques related to navigation of unmanned vehicles through unknown environments cluttered with obstacles, especially those that rigorously ensure collision avoidance (given certain assumptions about the system). This topic continues to be an active area of research, and we highlight some directions in which available approaches may be improved. The chapter discusses models of the sensors and vehicle kinematics, as well as assumptions about the environment and performance criteria. Methods applicable to stationary obstacles, moving obstacles, and scenarios with multiple vehicles are covered. In preference to global approaches based on full knowledge of the environment, particular attention is given to reactive methods based on only local sensory data, with a special focus on recently proposed navigation laws based on the sliding mode and model predictive control.

Chapter 4 deals with the problem of optimal navigation in the presence of obstacles. Specifically, a Dubins-car-like mobile robot travels among smooth and possibly non-convex obstacles, with a constraint on the curvature of their boundaries. The minimal-length path to a given steady target is explicitly found in the case where the scene is known in advance. Based on this solution, a novel reactive randomized algorithm of robot navigation in an unknown environment is proposed. It is proven that under this navigation law, the robot avoids collisions and reaches the steady target with probability 1. The performance of the algorithm is confirmed by computer simulations and outdoor experiments with a Pioneer P3-DX mobile wheeled robot.

Chapter 5 discusses the problem of reactively navigating a Dubins-car-like robot along an equidistant curve of an environmental object or, otherwise stated, following the boundary of this object with a predefined range margin. Two sensing scenarios are considered. In one of them, navigation is based on access to the distance to the nearest point of the boundary. The other scenario assumes that the distance to the boundary is measured perpendicularly to the robot’s centerline, and the angle of incidence of this perpendicular to the boundary is also accessible. This situation holds if, e.g., the measurements are supplied by range sensors rigidly mounted to the robot’s body at nearly right angles, or by a single sensor scanning a nearly perpendicular narrow sector. For both scenarios, reactive sliding mode control laws are proposed that drive the robot at a pre-specified distance from the boundary and thereafter maintain this distance. This is achieved without estimation of the boundary curvature and holds for boundaries with both convexities and concavities. Mathematically rigorous analysis of the proposed control laws is provided, including explicit account of the global geometry of the boundary. Computer simulations and experiments with real wheeled robots confirm the applicability and performance of the proposed guidance approach.

In Chapter 6, border-patrolling algorithms are applied to a more general problem of reactively navigating a nonholonomic Dubins-car like robot to a target through an environment cluttered with static obstacles. The proposed navigation strategy combines motions straight to the target, when possible, with bypassing obstacles in a close range with the aid of border patrolling algorithms, and also includes a set of rules regulating switches between these two regimes. Mathematically rigorous analysis of the navigation strategy is provided. Its efficiency is also illustrated by computer simulations.

In Chapter 7, we present a method for guidance of a Dubins-car-like mobile robot with saturated control toward a target in a steady simply connected maze-like environment. The robot always has access to the target relative bearing angle and the distance to the nearest point of the maze if it is within the given sensor range. The proposed control law is composed by biologically inspired reflex-level rules. Mathematically rigorous analysis of this law is provided, and its convergence and performance are confirmed by computer simulations and experiments with real robots.

Chapter 8 introduces and studies a simple biologically-inspired strategy for navigation a Dubins-car-like robot toward a target while avoiding collisions with moving obstacles. A mathematically rigorous analysis of the proposed approach is provided. The performance of the algorithm is demonstrated via experiments with real robots and extensive computer simulations.

Chapter 9 deals with a hard scenario where the environment of operation is cluttered with obstacles that may undergo arbitrary motions, including rotations and deformations, i.e., changes of shapes and sizes. A sliding mode based strategy for a unicycle-like robot navigation and guidance is presented. It is then applied to the problems of patrolling the border of a moving and deforming domain and reaching a target through a dynamic environment cluttered with moving obstacles. Mathematically rigorous analysis of the proposed approach is provided. The convergence and performance of the algorithm are demonstrated via experiments with real robots and extensive computer simulation.

Chapter 10 presents a novel reactive algorithm for collision free navigation of a nonholonomic robot in unknown complex dynamic environments with moving obstacles. The novelty is related to employment of an integrated representation of the information about the environment, which does not require separating obstacles and approximating their shapes by discs or polygons, and is very easy to obtain in practice. Moreover, the presented algorithm does not require any information on the obstacles’ velocities. Under the examined navigation algorithm, the robot efficiently seeks a short path through a crowd of moving or steady obstacles, rather than avoiding the crowd as a whole, like many other navigation algorithms do. A mathematically rigorous analysis of the proposed approach is provided. The performance of the algorithm is demonstrated via experiments with a real robot and extensive computer simulations.

Chapter 11 introduces and examines a novel purely reactive algorithm to navigate a planar mobile robot in densely cluttered environments with unpredictably moving and deforming obstacles. The algorithm is supplied with mathematically firm guarantees of not only avoidance of collisions with the obstacles but also achieving the global objective of perpetual drift in the desired direction. The efficiency of the method is illustrated by computer simulations.

While Chapters 4–11 deal with navigation of a single robot, Chapter 12 considers a multiple robot scenario. The objective is to simultaneously navigate several mobile robots with limited sensing and communication capabilities through an unknown static environment. In this chapter, we propose a decentralized, cooperative, reactive, model predictive control based collision avoidance scheme that plans short-range paths in the currently sensed part of the environment, and show that it is able to prevent collisions from occurring. The proposed scheme combines a main path-planning system with an auxiliary controller, which is employed to follow previously planned paths whenever the main system fails to update the path. Simulations and real-world testing in various scenarios confirm the method’s validity.

For the convenience of the reader, cross-references between chapters are kept to a minimum. Therefore, numbering of formulas and theorem-like objects (theorems, lemmas, assumptions, etc.) starts over with each new section and uses the section number followed by dot . and the object’s sequential number within the section. For example, Definition 2.1 refers to the first definition from the second section in the current chapter. In rare occasions when a reference to not the current chapter is, however, made, the chapter number is added to the beginning; for example, Assumption 3.4.2 refers to the second assumption in the fourth section from Chapter 3. The section number consists of two component parts separated by dot ., the figure before the dot refers to the chapter number and the figure after the dot refers to the position of the section within the chapter. Thus, Section 5.2 refers to the second section in Chapter 5. The following template is used for numbering subsections: the section number.the position of the subsection within the section. For example, Section 7.3.4 refers to the fourth subsection in Section 7.3. Since most of figures and tables are concentrated in the concluding sections of the chapters, the figures and tables are numbered per chapter, such as Fig. 4.3 referring to the third figure in Chapter 4. Unlike numbering of other items, equation numbers appear within parentheses, e.g., (3.8).

1.3 Sliding mode control

Due to the well-known benefits such as high insensitivity to noises and disturbance rejection feature, robustness against uncertainties and unmodeled dynamics, good dynamic response, good transient performance, reduced complexity of design and tuning, capacity to decouple a highly dimensional problem into a set of lower dimensional ones, ease of implementation, etc. [70], the sliding mode approach attracts a steady stream of interest in the area of motion control (we refer the reader to [71, 72] for detailed surveys). In particular, the origins of sliding mode based motion control with obstacle avoidance can be traced back to early 1990s [73, 74]. The listed benefits of the sliding mode approach is a particular concern for wheeled mobile robots, which are intrinsically subjected to various uncertainties, including, but not limited to, parametric disturbances, such as changes in mass, inertia, and friction, uncertainties of an inhomogeneous terrain, such as unexpected slippery conditions, upward and downward ramps, rough surfaces, tire sliding on ice and rolling on sand, uncertainties of aerodynamic forces, uncertainties in actuators and sensors, e.g., in magnifications factors, focal length, and orientation of a video-camera, etc. Furthermore highly complex, coupled, and nonlinear dynamics of typical wheeled robots entail that the control system design is commonly based on account of not the entire dynamics, leaving noticeable discrepancies between the model and the real dynamics of the robot. Further, the need for precise control and high performance is ever increasing in real-world applications.

The technical part of this book systematically applies the sliding mode approach to various ever open problems in the area of navigation, guidance, and control of wheeled mobile robots, though the status of the concerned sliding mode controller may vary from primary down to secondary between chapters. General issues on design and analysis of sliding mode controllers are well documented in numerous monographs and textbooks; see, e.g., [70, 75–80] and literature therein. For convenience of the reader, basic technicalities of sliding mode control will be recalled in Chapter 2 to the extent needed in this book.

The major obstacle to implementation of sliding mode controllers is a harmful phenomenon called chattering, i.e., undesirable finite frequency oscillations around the ideal trajectory due to unmodeled system dynamics and constraints. The problem of chattering elimination and reduction has extensive literature; see, e.g., [78, 81–90] and literature therein. It offers a variety of effective approaches, including continuous approximation of the discontinuity, inserting low-pass filters/observers into the control loop, combining sliding mode and adaptive control or fuzzy logic techniques, higher order sliding modes, etc.

In the research presented in this book, we judged the occurrence of chattering mostly at the stage of real-world experiments. Continuous approximation of the discontinuous control law in a boundary layer [70, 86] was used to eliminate chattering in some cases and as a prophylactic measure against chattering in other cases, whereas there were cases where anti-chattering treatment was not needed. These outcomes are specific to experimental platforms employed in our research. Whether chattering be encountered when implementing the proposed control laws on different platforms, it can be handled in accordance with the treatment recipes offered in the afore-mentioned literature.

1.4 Experimental equipment

In experimental verification of the algorithms developed in this book, the following three types of wheeled mobile robots were used:

• laboratorial wheeled robots Pioneer P3-DX;

• an intelligent autonomous wheelchair system;

• an autonomous hospital bed system.

1.4.1 Laboratorial wheeled robot Pioneer P3-DX

Most of the experiments presented in the book were performed with laboratorial wheeled robots Pioneer P3-DX (abbreviated as P3-DX) (Fig. 1.1). This differential drive robot from ActivMedia Robotics is intended for classroom and laboratory use. Its key