| Titre : | Toward Entertainment in Swarm Robotics : A Focus on Artistic Dynamic Patterns Transformation |
| Auteurs : | Belkacem Khaldi, Auteur ; Foudil Cherif, Directeur de thèse |
| Type de document : | Thése doctorat |
| Editeur : | Biskra [Algérie] : Faculté des Sciences Exactes et des Sciences de la Nature et de la Vie, Université Mohamed Khider, 2018 |
| Format : | 1 vol. (140 p.) / 30 cm |
| Langues: | Anglais |
| Mots-clés: | Swarm Robotics,Pattern Formation,Self-Organized Aggregating Pat- terns,Virtual Viscoelastic Model,Exogenous fault detection. |
| Résumé : |
In the last decade, a wide range of successful applications has been established to meet the increasing need for robot swarms in our daily lives. Bringing such technology into entertainment and artistic activities such as painting, music or even dancing, is a new challenge that will increase and intensify research efforts in this area to target a wide range of stakeholders. Inspired by this challenge, we aim to enrich existing researches in the field of robotics for entertainment, by studying how a large number of robots can be used in ceremonies. We believe that this can be achieved by focusing on a typical study of multi-robot systems, namely patterns formation. The latter, which is an important phenomena found in living organisms (e.g., animals and plants) and physical organisms (e.g., sand dunes or galaxies), is a challenge aimed at confronting the problem of organizing a group of robots in global formations or patterns. These formations could be either simple patterns such as circles, lines, uniform distribution within a circle or square, etc., or complex patterns consist of simple patterns. In this thesis, we are interested in designing and synthesizing controllers for robotics swarm systems to achieve patterns in a self-organized manner. Our approach is taken from the inspiration of nature, especially from the biomechanical forces involved in the studies of the inner cells on the one hand, and from the topological metric revealed in studies of bird flocks on the other hand. In order to produce self-organized aggregating patterns with robots swarm in an effective manner, we have devised many experimental ARGoS-based simulations (Autonomous Robots Go Swarming simulator) that allow us to study multiple aspects of self-organized collective behaviors. One of the main problems we focus on to study such behaviors using swarms of robots includes models of formation control, models of self-organized aggregating patterns, and fault detection in swarm robots formation control models. |
| Sommaire : |
Abstract iii Résumé v Acknowledgements vii 1 Introduction 1 1.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Motivation & Main Objectives . . . . . . . . . . . . . . . . . . . . . . . . 3 1.2.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.2.2 Main Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 1.3 Contributions and Related Publications . . . . . . . . . . . . . . . . . . 4 1.3.1 Preview of Contributions . . . . . . . . . . . . . . . . . . . . . . 4 1.3.2 Related Publications . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.4 Dissertation layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 I Background and Related Works 11 2 An Overview of Swarm Robotics 13 2.1 Swarm intelligence (SI) - an inspiration of Natural Swarm Systems . . 13 2.1.1 The Genius of Natural Swarm Systems . . . . . . . . . . . . . . 13 Bird flocking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Ants’ colonies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Swarming of bees . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Fish schools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.1.2 Swarm Intelligence Systems: Definition and Properties . . . . . 16 2.1.3 Natural Swarm Behavior based Meta-Heuristics Algorithms . . 18 2.2 Swarm Robotics – Swarm Intelligence applied to Multi-robot systems . 18 2.2.1 Multi-robotics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.2.2 Swarm Robotics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 2.2.3 Potential Application of Swarm Robotics . . . . . . . . . . . . . 21 2.2.4 Swarm Robotics Problems Focus . . . . . . . . . . . . . . . . . . 22 2.2.5 Involved projects and simulations . . . . . . . . . . . . . . . . . 24 Swarm robotics involved projects . . . . . . . . . . . . . . . . . 24 Swarm robotics simulation platforms . . . . . . . . . . . . . . . 24 x 2.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 3 Studies in Aggregating Patterns and Faults Detection within Swarm Robotics systems 29 3.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 3.2 Aggregation Patterns in Nature . . . . . . . . . . . . . . . . . . . . . . . 29 3.2.1 Aggregation in Nature . . . . . . . . . . . . . . . . . . . . . . . . 29 3.2.2 Emergence of Patterns in Aggregating Natural Swarms . . . . . 30 3.2.3 Case study: Flocking Patterns . . . . . . . . . . . . . . . . . . . . 30 Models on Flocking Patterns . . . . . . . . . . . . . . . . . . . . 31 3.3 Aggregation Patterns in Swarm Robotics . . . . . . . . . . . . . . . . . 32 3.3.1 Aggregation in Swarm Robotics . . . . . . . . . . . . . . . . . . 32 3.3.2 Emergence of Patterns in Aggregating Robotics Swarm . . . . . 32 3.4 Approaches on Swarm Robotics Aggregation Patterns . . . . . . . . . . 33 3.4.1 Cue-Based Methods . . . . . . . . . . . . . . . . . . . . . . . . . 33 3.4.2 Self-organized based Methods . . . . . . . . . . . . . . . . . . . 34 Probabilistic Approach . . . . . . . . . . . . . . . . . . . . . . . . 34 Deterministic Approach . . . . . . . . . . . . . . . . . . . . . . . 35 Artificial Evolution Approach . . . . . . . . . . . . . . . . . . . . 35 Morphogenesis Inspired Approach . . . . . . . . . . . . . . . . . 35 Artificial Physics Approach . . . . . . . . . . . . . . . . . . . . . 36 3.5 Detecting Faulty Robots in Swarm Robotics Systems . . . . . . . . . . . 37 3.5.1 Faults in Engineered Systems . . . . . . . . . . . . . . . . . . . . 37 3.5.2 Faults in Swarm Robotics Systems . . . . . . . . . . . . . . . . . 37 Fault Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38 3.5.3 Fault Detection in Engineered Systems . . . . . . . . . . . . . . 39 Legacy Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Model-based Approaches . . . . . . . . . . . . . . . . . . . . . . 40 Data-driven Approaches . . . . . . . . . . . . . . . . . . . . . . . 40 3.5.4 Fault Detection in Robotics Swarm Systems . . . . . . . . . . . . 41 Endogenous Approaches . . . . . . . . . . . . . . . . . . . . . . 41 Exogenous Approaches . . . . . . . . . . . . . . . . . . . . . . . 42 3.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 II Self-Organized Patterns in Aggregating robots swarm 45 4 Material and Methods 47 4.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 4.2 Simulation platform (ARGoS) . . . . . . . . . . . . . . . . . . . . . . . . 47 4.3 Robotic Platform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 4.3.1 Foot-bot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 4.3.2 e-puck . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 xi 4.4 On-board Sensing/Actuating System . . . . . . . . . . . . . . . . . . . . 51 4.4.1 Infrared range and bearing sensors . . . . . . . . . . . . . . . . . 52 4.4.2 Infrared proximity sensors . . . . . . . . . . . . . . . . . . . . . 53 4.4.3 Wheels Actuator . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 4.5 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 4.5.1 The Viscoelastic Model: A Bio-mechanic inspired Model . . . . 56 Overview of the inner cell Bio-mechanics . . . . . . . . . . . . . 56 The swarm robots viscoelastic interaction model . . . . . . . . . 56 4.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 5 Basic Geometric Formations 61 5.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 5.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 5.2.1 The Task & the Experimental Setup . . . . . . . . . . . . . . . . 61 5.2.2 Robot’ Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Proximity Control (PC) . . . . . . . . . . . . . . . . . . . . . . . . 62 Repulsive Control (RC) . . . . . . . . . . . . . . . . . . . . . . . 63 Forward Dependent Angular Motion Control (FDAMC) . . . . 64 5.3 Algorithms & Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 5.3.1 Regular Shapes Formation within foot-bots . . . . . . . . . . . . 65 5.3.2 Circle Formation within e-pucks . . . . . . . . . . . . . . . . . . 68 Circumscribed Circle Theory . . . . . . . . . . . . . . . . . . . . 69 Circle Formation via the Virtual Viscoelastic Control Model . . 69 5.4 Performance Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 5.4.1 Performance Metrics . . . . . . . . . . . . . . . . . . . . . . . . . 72 Group Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Mean Distance Error . . . . . . . . . . . . . . . . . . . . . . . . . 72 5.4.2 Results within the Metrics . . . . . . . . . . . . . . . . . . . . . . 72 5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 6 Topological Approaches in Self-organized Aggregating Patterns 77 6.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 6.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 6.2.1 The Task & the Experimental Setup . . . . . . . . . . . . . . . . 77 6.2.2 Robot’ Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 The K-NN Topological Approach . . . . . . . . . . . . . . . . . . 78 The DW-KNN Topological Approach . . . . . . . . . . . . . . . 79 6.3 Results & Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 6.3.1 Self-organized Aggregation Patterns basing on the KNN approach. .. . . . . . 82 6.3.2 Self-organized Aggregation Patterns basing on the DW-KNN approach ........ . 83 6.4 Performance Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . ... . 86 6.4.1 Performance Metrics . . . . . . . . . . . . . . . . . . . . . . . . . 87 Distance-Weighted Distribution Quality . . . . . . . . . . . . . . 87 Aggregation Quality . . . . . . . . . . . . . . . . . . . . . . . . . 88 Dispersion Quality . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Averaged Mean Distance Error . . . . . . . . . . . . . . . . . . . 89 6.4.2 Analysis in Normal Circumstances . . . . . . . . . . . . . . . . . 89 6.4.3 The Effect of Sensory Noise on the Performnce of the DW-KNN Approch . . . . . 90 6.5 Analysis of the Two Approaches within Different Noise Models . . . . 93 6.5.1 Effect of Uniform Noise on the Performance Proposed Approach. .. . . . . . . . . . . . . . . . . . . . 93 6.5.2 Effect of Gaussian Noise on the Performance Proposed Approach 94 6.5.3 Effect of Mean Shift Noise on the Performance Proposed Approach . . . . . . . . . .. . . . . . . . 95 6.6 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 7 Detecting Faulty Robots in Aggregating robots Swarms 99 7.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 7.2 the task and the main objective . . . . . . . . . . . . . . . . . . . . . . . 99 7.3 PCA-based monitoring approaches . . . . . . . . . . . . . . . . . . . . . 100 7.3.1 Feature extraction using PCA . . . . . . . . . . . . . . . . . . . . 100 7.3.2 PCA-based fault detection . . . . . . . . . . . . . . . . . . . . . . 101 7.4 Univariate statistical control charts . . . . . . . . . . . . . . . . . . . . . 102 7.4.1 EWMA monitoring charts . . . . . . . . . . . . . . . . . . . . . . 102 7.4.2 Cumulative sum (CUSUM) charts . . . . . . . . . . . . . . . . . 103 7.4.3 Combining PCA with CUSUM and EWMA charts . . . . . . . 104 7.5 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 7.5.1 PCA modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 7.5.2 Detection results . . . . . . . . . . . . . . . . . . . . . . . . . . . 107 Case (A): Abrupt fault detection . . . . . . . . . . . . . . . . . . 108 Case (B): Intermittent fault . . . . . . . . . . . . . . . . . . . . . 109 Case (C): Random walk fault . . . . . . . . . . . . . . . . . . . . 110 Case (D): Complete stop fault . . . . . . . . . . . . . . . . . . . . 110 Case (E): Drift failure detection . . . . . . . . . . . . . . . . . . . 112 7.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 8 Conclusion and FutureWorks 115 8.1 FutureWorks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 A Algorithms and Listings 119 A.1 An Example of an XML based ARGoS Configuration File . . . . . . . . 119 A.2 The Overall KNN Control Algorithm Implemented in a Foot-Bot Robot 122 A.3 The Overall DW-KNN Control Algorithm Implemented in a Foot-Bot Robot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125 Bibliography 127 |
| En ligne : | http://thesis.univ-biskra.dz/3859/1/01_thesis_khaldi.pdf |
Disponibilité (1)
| Cote | Support | Localisation | Statut |
|---|---|---|---|
| TINF/131 | Théses de doctorat | bibliothèque sciences exactes | Consultable |
