No description available.

<div>Collision avoidance design plays an essential role in autonomous vehicle technology. It's an attractive research area that will need much experimentation in the future. This research area is very important for providing the maximum safety to automated vehicles, which have to be tested several times under different circumstances for safety before use in real life. This thesis proposes a method for designing and presenting a collision avoidance maneuver by using a model predictive controller with a moving obstacle for automated vehicles. It consists of a plant model, an adaptive MPC controller, and a reference trajectory. The proposed strategy applies a dynamic bicycle model as the plant model, adaptive model predictive controller for the lateral control, and a custom reference trajectory for the scenario design. The model was developed using the Model Predictive Control Toolbox and Automated Driving Toolbox in Matlab. Builtin tools available in Matlab/Simulink were used to verify the modeling approach and analyze the performance of the system. The major contribution of this thesis work was implementing a novel dynamic obstacle avoidance control method for automated vehicles. The study used validated parameters obtained from previous research. The novelty of this research was performing the studies using a MPC based controller instead of a sliding mode controller, that was primarily used in other studies. The results obtained from the study are compared with the validated models. The comparisons consisted of the lateral overlap,lateral error, and steering angle simulation results between the models. Additionally,this study also included outcomes for the yaw angle. The comparisons and other outcomes obtained in this study indicated that the developed control model produced reasonably acceptable results and recommendations for future studies.</div>

In present road traffic system, drone-network based traffic monitoring using the Internet of Vehicles (IoVs) is a promising solution. However, camera-based traffic monitoring does not collect complete data, cover all areas, provide quick medical services, or take vehicle follow-ups in case of an incident. Drone-based system helps to derive important information (such as commuter's behavior, traffic patterns, vehicle follow-ups) and sends this information to centralized or distributed authorities for making traffic diversions or necessary decisions as per laws. The present approaches fail to meet the requirements such as (i) collision free, (ii) drone navigation, and (iii) less computational and communicational overheads. This work has considered the collision-free drone-based movement strategies for road traffic monitoring using Software Defined Networking (SDN). The SDN controllable drone network results in lesser overhead over drones and provide efficient drone-device management. In simulation, two case studies are simulated using JaamSim simulator. Results show that the zones-based strategy covers a large area in few hours and consume 5 kWs to 25 kWs energy for 150 drones (Case study 1). Zone-less based strategies (case study-2) show that the energy consumption lies between 5 kWs to 18 kWs for 150 drones. Further, the use of SDN-based drones controller reduces the overhead over drone-network and increases the area coverage with a minimum of 1.2% and maximum of 2.6%. Simulation (using AnyLogic simulator) shows the 3D view of successful implementation of collision free strategies.

京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第21918号 / 情博第701号 / 新制||情||120(附属図書館) / 京都大学大学院情報学研究科システム科学専攻 / (主査)准教授 西原 修, 教授 大塚 敏之, 教授 加納 学 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM

In present road traffic system, drone-network based traffic monitoring using the Internet of Vehicles (IoVs) is a promising solution. However, camera-based traffic monitoring does not collect complete data, cover all areas, provide quick medical services, or take vehicle follow-ups in case of an incident. Drone-based system helps to derive important information (such as commuter's behavior, traffic patterns, vehicle follow-ups) and sends this information to centralized or distributed authorities for making traffic diversions or necessary decisions as per laws. The present approaches fail to meet the requirements such as (i) collision free, (ii) drone navigation, and (iii) less computational and communicational overheads. This work has considered the collision-free drone-based movement strategies for road traffic monitoring using Software Defined Networking (SDN). The SDN controllable drone network results in lesser overhead over drones and provide efficient drone-device management. In simulation, two case studies are simulated using JaamSim simulator. Results show that the zones-based strategy covers a large area in few hours and consume 5 kWs to 25 kWs energy for 150 drones (Case study 1). Zone-less based strategies (case study-2) show that the energy consumption lies between 5 kWs to 18 kWs for 150 drones. Further, the use of SDN-based drones controller reduces the overhead over drone-network and increases the area coverage with a minimum of 1.2% and maximum of 2.6%. Simulation (using AnyLogic simulator) shows the 3D view of successful implementation of collision free strategies.

Indiana University-Purdue University Indianapolis (IUPUI) / Traffic accidents cause millions of injuries and tens of thousands of fatalities per year worldwide. This thesis briefly reviews different types of active safety systems designed to reduce the number of accidents. Focusing on lane departure, a leading cause of crashes involving fatalities, we examine a lane-keeping system proposed by Minoiu Enache et al.They proposed a switched linear feedback (LMI) controller and provided two switching laws, which limit driver torque and displacement of the front wheels from the center of the lane. In this thesis, a state feedback (LQR) controller has been designed. Also, a new switching logic has been proposed which is based on driver's torque, lateral offset of the vehicle from the center of the lane and relative yaw angle. The controller activates assistance torque when the driver is deemed inattentive. It is deactivated when the driver regains control. Matlab/Simulink modeling and simulation environment is used to verify the results of the controller. In comparison to the earlier switching strategies, the maximum values of the state variables lie very close to the set of bounds for normal driving zone. Also, analysis of the controller’s root locus shows an improvement in the damping factor, implying better system response.

A quadrotor is an omnidirectional unmanned air vehicle that is suitable for indoor flight because of its ability to hover and maneuver in confined spaces. The remote operation of this type of vehicle is difficult due to a lack of sensory perception; typically, the view from the onboard camera is the only information transmitted to the pilot. This thesis proposes using force feedback exerted by the command input device on the hand of the pilot to assist in avoiding collisions while navigating in indoor environments. Five candidate algorithms are presented for calculating the forces to be felt by the pilot based on the quadrotor's position and velocity in the indoor environment. The candidates include a parametric algorithm based on the dynamics of the quadrotor, two time-to-impact algorithms, and two algorithms that employ virtual springs between the quadrotor and obstacles. A method of incorporating the position of the command input device to improve the usability and effectiveness of the algorithms is also presented. A framework for simulating the quadrotor dynamics, indoor environment, and force feedback algorithms is described. In the simulation, the pilot commands a simulated quadrotor, using a commercial haptic interface, as it flies in an indoor environment. The pilot receives force feedback cues as the quadrotor navigates around obstacles. Two methods of control were used for the simulation. In the first method, displacements of the haptic interface correspond to velocity commands to the quadrotor. In the second method, displacements of the input correspond to desired roll and pitch commands. Two user study experiments, one for each control method, were performed to compare the force feedback algorithms in simulation. The results of the velocity control experiment suggest that higher force levels help to avoid collisions and that the time to impact algorithm results in fewer collisions than having no force, but is not significantly better than the other algorithms. The results of the angle control experiment suggest that the time to impact algorithm is clearly the best in terms of hits and hit length and has no disadvantages compared to the other algorithms. Finally, to demonstrate the force feedback algorithms and software in a real-world environment, the system was interfaced with a physical quadrotor. The quadrotor system is described and the results of the tests are presented.

Unmanned vehicle systems, specifically Unmanned Air Vehicles (UAVs) and Unmanned Ground Vehicles (UGVs) have found potential use in both military and civilian applications. For many applications, unmanned vehicle systems are required to navigate in urban environments where obstacles with various types and sizes exist. The main contribution of this research is to offer vision-based path planning, collision avoidance, and target tracking strategies for Unmanned Air and Ground vehicles operating in urban environments. Two vision-based local-level frame mapping and planning techniques are first developed for Miniature Air Vehicles (MAVs). The techniques build maps and plan paths in the local-level frame of MAVs directly using the camera measurements without transforming to the inertial frame. Using a depth map of an environment obtained by computer vision methods, the first technique employs an extended Kalman Filter (EKF) to estimate the range, azimuth to, and height of obstacles, and constructs local spherical maps around MAVs. Based on the maps, the Rapidly-Exploring Random Tree (RRT) algorithm is used to plan collision-free Dubins paths. The second technique constructs local multi-resolution maps using an occupancy grid, which give higher resolution to the areas that are close to MAVs and give lower resolution to the areas that are far away. The maps are built using a log-polar representation. The two planning techniques are demonstrated in simulation and flight tests. Based on the observation that a camera does not provide accurate time-to-collision (TTC) measurements, two and three dimensional observability-based planning algorithms are explored. The techniques estimate both TTC and bearing using bearing-only measurements. A nonlinear observability analysis of state estimation process is conducted to obtain the conditions for complete observability of the system. Using the conditions, the observability-based planning algorithms are designed to minimize the estimation uncertainties while simultaneously avoiding collisions. The two dimensional planning algorithm parameterizes an obstacle using TTC and azimuth, and constructs local polar maps. The three dimensional planning algorithm parameterizes an obstacle using inverse TTC, azimuth, and elevation, and constructs local spherical maps. The algorithms are demonstrated in simulation. Lastly, a probabilistic path planning algorithm is developed for tracking a moving target in urban environments using UAVs and UGVs. The algorithm takes into account occlusions due to obstacles. It models the target using a dynamic occupancy grid and updates the target location using a Bayesian filter. Based on the target's current and probable future locations, a decentralized path planning algorithm is designed to generate suboptimal paths that maximize the sum of the joint probability of detection for all vehicles over a finite look-ahead horizon. Results demonstrate the planning algorithm is successful in solving the moving target tracking problem in urban environments.

Unmanned aircraft systems (UAS) represent the future of modern aviation. Over the past 10 years their use abroad by the military has become commonplace for surveillance and combat. Unfortunately, their use at home has been far more restrictive. Due to safety and regulatory concerns, UAS are prohibited from flying in the National Airspace System without special authorization from the FAA. One main reason for this is the lack of an on-board pilot to "see and avoid" other air traffic and thereby maintain the safety of the skies. Development of a comparable capability, known as "Sense and Avoid" (SAA), has therefore become a major area of focus. This research focuses on the SAA problem as it applies specifically to small UAS. Given the size, weight, and power constraints on these aircraft, current approaches fail to provide a viable option. To aid in the development of a SAA system for small UAS, various simulation and hardware tools are discussed. The modifications to the MAGICC Lab's simulation environment to provide support for multiple agents is outlined. The use of C-MEX s-Functions to improve simulation performance and code portability is also presented. For hardware tests, two RC airframes were constructed and retrofitted with autopilots to allow autonomous flight. The development of a program to interface with the ground control software and run the collision avoidance algorithms is discussed as well. Intruder sensing is accomplished using a low-power, low-resolution radar for detection and an Extended Kalman Filter (EKF) for tracking. The radar provides good measurements for range and closing speed, but bearing measurements are poor due to the low-resolution. A novel method for improving the bearing approximation using the raw radar returns is developed and tested. A four-state EKF used to track the intruder's position and trajectory is derived and used to provide estimates to the collision avoidance planner. Simulation results and results from flight tests using a simulated radar are both presented. To effectively plan collision avoidance paths a tree-branching path planner is developed. Techniques for predicting the intruder position and creating safe, collision-free paths using the estimates provided by the EKF are presented. A method for calculating the cost of flying each path is developed to allow the selection of the best candidate path. As multiple duplicate paths can be created using the branching planner, a strategy to remove these paths and greatly increase computation speed is discussed. Both simulation and hardware results are presented for validation.

Today a group of automated guided vehicles at Toyota Material Handling Manufacturing Sweden detect and avoid objects primarily by using 2D-LiDAR, with shortcomings being the limitation of only scanning the area in a 2D plane and missing objects close to the ground. Several dynamic obstacles exist in the environment of the vehicles. Protruding forks are one such obstacle, impossible to detect and avoid with the current choice of sensor and its placement. This thesis investigates possible solutions and limitations of using a single RGB camera for obstacle detection, tracking, and avoidance. The obstacle detection uses the deep learning model YOLOv5s. A solution for semi-automatic data gathering and labeling is designed, and pre-trained weights are chosen to minimize the amount of labeled data needed. Two different approaches are implemented for the tracking of the object. The YOLOv5s detection is the foundation of the first, where 2D-bounding boxes are used as measurements in an Extended Kalman Filter (EKF). Fiducial markers build up the second approach, used as measurements in another EKF. A state lattice motion planner is designed to find a feasible path around the detected obstacle. The chosen graph search algorithm is ARA*, designed to initially find a suboptimal path and improve it if time allows.  The detection works successfully with an average precision of 0.714. The filter using 2D-bounding boxes can not differentiate between a clockwise and counterclockwise rotation, but the performance is improved when a measurement of rotation is included. Using ARA* in the motion planner, the solution sufficiently avoids the obstacles.