[Springer Theses] Decision-making Strategies for Automated Driving in Urban Environments ||

"This book describes an effective decision-making and planning architecture for enhancing the navigation capabilities of automated vehicles in the presence of non-detailed, open-source maps. The system involves dynamically obtaining road corridors from map information and utilizing a camera-based lane detection system to update and enhance the navigable space in order to address the issues of intrinsic uncertainty and low-fidelity. An efficient and human-like local planner then determines, within a probabilistic framework, a safe motion trajectory, ensuring the continuity of the path curvature and limiting longitudinal and lateral accelerations. LiDAR-based perception is then used to identify the driving scenario, and subsequently re-plan the trajectory, leading in some cases to adjustment of the high-level route to reach the given destination. The method has been validated through extensive theoretical and experimental analyses, which are reported here in detail."-- Provided by publisher Supervisor’s Foreword 7 Parts of this thesis have been published in the following articles: 8 Journal Papers 8 Book Chapters 8 Conferences 8 Datasets in Online Data Repository 9 Acknowledgements 10 Contents 12 Abbreviations 15 1 Introduction 17 1.1 Overview 17 1.2 Motivation and Framework 18 1.3 Objectives 20 1.4 Thesis Outline 21 References 22 2 Literature Overview 24 2.1 Introduction 24 2.2 Map Generation 29 2.2.1 Smooth Road Geometry Models 30 2.2.2 Map-Based ADAS 31 2.3 Intention Prediction, Risk Estimation and Decision-Making 32 2.3.1 Reward-Based Algorithms for Motion Planning 32 2.3.2 Risk Based Motion Planning in Dynamic and Uncertain Environments 33 2.4 Motion Planning 35 2.4.1 Path Primitives 37 2.4.2 Considering Localization Uncertainty in Motion Planning 37 References 38 3 Decision-Making Architecture 43 3.1 Introduction 43 3.2 Prior State of the Architecture 43 3.2.1 Mission and Planner Modules 44 3.2.2 Control Module 46 3.2.3 Prior Architecture Remarks 46 3.3 Contributions to the Architecture 48 3.3.1 Global Planning Capabilities 49 3.3.2 Local Planning Capabilities 50 3.3.3 Additional Capabilities 50 References 51 4 Global Planning and Mapping 52 4.1 Introduction 52 4.2 Assumptions 53 4.3 Global Planner 54 4.4 Road Corridor Generation from Low Fidelity Maps 56 4.4.1 Road Corridor Generation Algorithm Overview 57 4.4.2 Bézier Adjustment 60 4.4.3 Corridor Generation 63 4.4.4 Validation and Results of the Road Corridor Generation Algorithm 64 4.5 Vision-Based Road Corridor Adaptation 66 4.5.1 Image Processing for Lane Detection 67 4.5.2 Mapping and Validity Checking 71 4.6 Considering Localization Uncertainty When Using Road Corridors 74 References 80 5 Motion Prediction and Manoeuvre Planning 82 5.1 Introduction 82 5.2 Assumptions 83 5.3 Risk Estimation in Urban Environments 84 5.3.1 Dynamic Bayesian Network Model 84 5.3.2 Particle Filter 87 5.3.3 Results 89 5.4 Simplified Motion Prediction for Dynamic Objects 99 5.5 Manoeuvre Planner 100 References 102 6 Optimal Trajectory Generation 103 6.1 Introduction 103 6.2 Assumptions 104 6.3 Optimal Path Planning 105 6.3.1 Problem Statement 105 6.3.2 Comparison Framework Description 118 6.3.3 Experiments and Results 121 6.3.4 Comparative Conclusions 130 6.4 Trajectory Generation 130 6.4.1 Choosing the Planning Approach 130 6.4.2 Collision Checking 133 6.4.3 Speed Profile Generation 134 6.4.4 Trajectory Generation Algorithm 137 6.4.5 Trajectory Generation Results 142 6.5 Occupancy Grid-Based Motion Planning 149 6.5.1 Candidates Evaluation with Occupancy Grid 150 6.5.2 Motion Planning Results Using the Occupancy Grid 151 References 162 7 Integration and Demonstrations 164 7.1 Introduction 164 7.2 Experimental Platform 164 7.2.1 Experimental Platform Components 164 7.2.2 Software Architecture 173 7.2.3 Implementation and Integration of New Software Modules in the Architecture 175 7.3 Demonstrations 190 7.3.1 Demonstration at IROS 2018 190 7.3.2 Demonstration at S-Moving 2018 194 References 197 8 Conclusions 199 8.1 General Conclusions 199 8.1.1 Contributions 200 8.2 Future Work 201 8.3 Dissemination 202 Appendix Author Biography 205