Showing papers on "Collision avoidance published in 2017"

Path Planning and Tracking for Vehicle Collision Avoidance Based on Model Predictive Control With Multiconstraints

Abstract: A path planning and tracking framework is presented to maintain a collision-free path for autonomous vehicles. For path-planning approaches, a 3-D virtual dangerous potential field is constructed as a superposition of trigonometric functions of the road and the exponential function of obstacles, which can generate a desired trajectory for collision avoidance when a vehicle collision with obstacles is likely to happen. Next, to track the planned trajectory for collision avoidance maneuvers, the path-tracking controller formulated the tracking task as a multiconstrained model predictive control (MMPC) problem and calculated the front steering angle to prevent the vehicle from colliding with a moving obstacle vehicle. Simulink and CarSim simulations are conducted in the case where moving obstacles exist. The simulation results show that the proposed path-planning approach is effective for many driving scenarios, and the MMPC-based path-tracking controller provides dynamic tracking performance and maintains good maneuverability.

Decentralized non-communicating multiagent collision avoidance with deep reinforcement learning

Abstract: Finding feasible, collision-free paths for multiagent systems can be challenging, particularly in non-communicating scenarios where each agent's intent (e.g. goal) is unobservable to the others. In particular, finding time efficient paths often requires anticipating interaction with neighboring agents, the process of which can be computationally prohibitive. This work presents a decentralized multiagent collision avoidance algorithm based on a novel application of deep reinforcement learning, which effectively offloads the online computation (for predicting interaction patterns) to an offline learning procedure. Specifically, the proposed approach develops a value network that encodes the estimated time to the goal given an agent's joint configuration (positions and velocities) with its neighbors. Use of the value network not only admits efficient (i.e., real-time implementable) queries for finding a collision-free velocity vector, but also considers the uncertainty in the other agents' motion. Simulation results show more than 26% improvement in paths quality (i.e., time to reach the goal) when compared with optimal reciprocal collision avoidance (ORCA), a state-of-the-art collision avoidance strategy.

Collision Avoidance and Stabilization for Autonomous Vehicles in Emergency Scenarios

Abstract: Emergency scenarios may necessitate autonomous vehicle maneuvers up to their handling limits in order to avoid collisions. In these scenarios, vehicle stabilization becomes important to ensure that the vehicle does not lose control. However, stabilization actions may conflict with those necessary for collision avoidance, potentially leading to a collision. This paper presents a new control structure that integrates path tracking, vehicle stabilization, and collision avoidance and mediates among these sometimes conflicting objectives by prioritizing collision avoidance. It can even temporarily violate vehicle stabilization criteria if needed to avoid a collision. The framework is implemented using model predictive and feedback controllers. Incorporating tire nonlinearities into the model allows the controller to use all of the vehicle’s performance capability to meet the objectives. A prediction horizon comprised of variable length time steps integrates the different time scales associated with stabilization and collision avoidance. Experimental data from an autonomous vehicle demonstrate the controller safely driving at the vehicle’s handling limits and avoiding an obstacle suddenly introduced in the middle of a turn.

Uncertainty-Aware Reinforcement Learning for Collision Avoidance

Abstract: Reinforcement learning can enable complex, adaptive behavior to be learned automatically for autonomous robotic platforms. However, practical deployment of reinforcement learning methods must contend with the fact that the training process itself can be unsafe for the robot. In this paper, we consider the specific case of a mobile robot learning to navigate an a priori unknown environment while avoiding collisions. In order to learn collision avoidance, the robot must experience collisions at training time. However, high-speed collisions, even at training time, could damage the robot. A successful learning method must therefore proceed cautiously, experiencing only low-speed collisions until it gains confidence. To this end, we present an uncertainty-aware model-based learning algorithm that estimates the probability of collision together with a statistical estimate of uncertainty. By formulating an uncertainty-dependent cost function, we show that the algorithm naturally chooses to proceed cautiously in unfamiliar environments, and increases the velocity of the robot in settings where it has high confidence. Our predictive model is based on bootstrapped neural networks using dropout, allowing it to process raw sensory inputs from high-bandwidth sensors such as cameras. Our experimental evaluation demonstrates that our method effectively minimizes dangerous collisions at training time in an obstacle avoidance task for a simulated and real-world quadrotor, and a real-world RC car. Videos of the experiments can be found at this https URL.

Sampling-Based Path Planning for UAV Collision Avoidance

Abstract: The ability to avoid collisions with moving obstacles, such as commercial aircraft is critical to the safe operation of unmanned aerial vehicles (UAVs) and other air traffic. This paper presents the design and implementation of sampling-based path planning methods for a UAV to avoid collision with commercial aircraft and other moving obstacles. In detail, the authors develop and demonstrate a method based on the closed-loop rapidly-exploring random tree algorithm and three variations of it. The variations are: 1) simplification of trajectory generation strategy; 2) utilization of intermediate waypoints; 3) collision prediction using reachable set. The methods were validated in software-in-the-loop simulations, hardware-in-the-loop simulations, and real flight experiments. It is shown that the algorithms are able to generate collision free paths in real time for the different types of UAVs among moving obstacles of different numbers, approaching angles, and speeds.

Collision Avoidance for Cooperative UAVs With Optimized Artificial Potential Field Algorithm

Abstract: Unmanned aerial vehicle (UAV) systems are one of the most rapidly developing, highest level and most practical applied unmanned aerial systems. Collision avoidance and trajectory planning are the core areas of any UAV system. However, there are theoretical and practical problems associated with the existing methods. To manage these problems, this paper presents an optimized artificial potential field (APF) algorithm for multi-UAV operation in 3-D dynamic space. The classic APF algorithm is restricted to single UAV trajectory planning and usually fails to guarantee the avoidance of collisions. To overcome this challenge, a method is proposed with a distance factor and jump strategy to solve common problems, such as unreachable targets, and ensure that the UAV will not collide with any obstacles. The method considers the UAV companions as dynamic obstacles to realize collaborative trajectory planning. Furthermore, the jitter problem is solved using the dynamic step adjustment method. Several resolution scenarios are illustrated. The method has been validated in quantitative test simulation models and satisfactory results were obtained in a simulated urban environment.

Fast, On-line Collision Avoidance for Dynamic Vehicles Using Buffered Voronoi Cells

Abstract: This letter presents a distributed collision avoidance algorithm for multiple dynamic vehicles moving in arbitrary dimensions. In our algorithm, each robot continually computes its buffered Voronoi cell (BVC) and plans its path within the BVC in a receding horizon fashion. We prove that our algorithm guarantees collision avoidance for robots with single integrator dynamics. We show that our algorithm has computational complexity of $O(k)$ , which is the same as that of the optimal reciprocal collision avoidance (ORCA) algorithm, and is considerably faster than model predictive control (MPC) and sequential convex programming (SCP) based approaches. Moreover, ORCA and MPC-SCP require relative position, velocity, and even other information, to be exchanged over a communication network among the robots. Our algorithm only requires the sensed relative position, and therefore is well suited for on-line implementation as it does not require a communication network, and it works well with noisy relative position sensors. Furthermore, we provide an extension of our algorithm to robots with higher-order dynamics like quadrotors. We demonstrate the capabilities of our algorithm by comparing it to ORCA in multiple benchmark simulation scenarios, and we present results of over 70 experimental trials using five quadrotors in a motion capture environment.

Path planning for autonomous vehicles using model predictive control

Abstract: Path planning for autonomous vehicles in dynamic environments is an important but challenging problem, due to the constraints of vehicle dynamics and existence of surrounding vehicles. Typical trajectories of vehicles involve different modes of maneuvers, including lane keeping, lane change, ramp merging, and intersection crossing. There exist prior arts using the rule-based high-level decision making approaches to decide the mode switching. Instead of using explicit rules, we propose a unified path planning approach using Model Predictive Control (MPC), which automatically decides the mode of maneuvers. To ensure safety, we model surrounding vehicles as polygons and develop a type of constraints in MPC to enforce the collision avoidance between the ego vehicle and surrounding vehicles. To achieve comfortable and natural maneuvers, we include a lane-associated potential field in the objective function of the MPC. We have simulated the proposed method in different test scenarios and the results demonstrate the effectiveness of the proposed approach in automatically generating reasonable maneuvers while guaranteeing the safety of the autonomous vehicle.

Deep-Learned Collision Avoidance Policy for Distributed Multiagent Navigation

Abstract: High-speed, low-latency obstacle avoidance that is insensitive to sensor noise is essential for enabling multiple decentralized robots to function reliably in cluttered and dynamic environments. While other distributed multiagent collision avoidance systems exist, these systems require online geometric optimization where tedious parameter tuning and perfect sensing are necessary. We present a novel end-to-end framework to generate reactive collision avoidance policy for efficient distributed multiagent navigation. Our method formulates an agent's navigation strategy as a deep neural network mapping from the observed noisy sensor measurements to the agent's steering commands in terms of movement velocity. We train the network on a large number of frames of collision avoidance data collected by repeatedly running a multiagent simulator with different parameter settings. We validate the learned deep neural network policy in a set of simulated and real scenarios with noisy measurements and demonstrate that our method is able to generate a robust navigation strategy that is insensitive to imperfect sensing and works reliably in all situations. We also show that our method can be well generalized to scenarios that do not appear in our training data, including scenes with static obstacles and agents with different sizes. Videos are available at https://sites.google.com/view/deepmaca.

A Differential Game Approach to Multi-agent Collision Avoidance

Abstract: A multi-agent system consisting of $N$ agents is considered. The problem of steering each agent from its initial position to a desired goal while avoiding collisions with obstacles and other agents is studied. This problem, referred to as the multi-agent collision avoidance problem , is formulated as a differential game. Dynamic feedback strategies that approximate the feedback Nash equilibrium solutions of the differential game are constructed and it is shown that, provided certain assumptions are satisfied, these guarantee that the agents reach their targets while avoiding collisions.

Active collision avoidance for human–robot collaboration driven by vision sensors

Abstract: Establishing safe human–robot collaboration is an essential factor for improving efficiency and flexibility in today’s manufacturing environment. Targeting safety in human–robot collaboration, this paper reports a novel approach for effective online collision avoidance in an augmented environment, where virtual three-dimensional 3D models of robots and real images of human operators from depth cameras are used for monitoring and collision detection. A prototype system is developed and linked to industrial robot controllers for adaptive robot control, without the need of programming by the operators. The result of collision detection reveals four safety strategies: the system can alert an operator, stop a robot, move away the robot, or modify the robot’s trajectory away from an approaching operator. These strategies can be activated based on the operator’s existence and location with respect to the robot. The case study of the research further discusses the possibility of implementing the developed method in realistic applications, for example, collaboration between robots and humans in an assembly line.

Realistic car-following models for microscopic simulation of adaptive and cooperative adaptive cruise control vehicles

Abstract: Adaptive cruise control (ACC) and cooperative adaptive cruise control (CACC) are important technologies for the achievement of vehicle automation, and their effect on traffic systems generally is evaluated with microscopic traffic simulations. A successful simulation requires realistic vehicle behavior and minimal vehicle collisions. However, most existing ACC-CACC simulation studies used simplified models that were not based on real vehicle response. The studies rarely addressed collision avoidance in the simulation. The study presented in this paper developed a realistic and collision-free car-following model for ACC-CACC vehicles. A multiregime model combining a realistic ACC-CACC system with driver intervention for vehicle longitudinal motions is proposed. This model assumes that a human driver resumes vehicle control either according to his or her assessment or after a collision warning asks the driver to take over. The proposed model was tested in a wide range of scenarios to explore model performance and collision possibilities. The testing scenarios included three regular scenarios of stop-and-go, approaching, and cut-out maneuvers, as well as two extreme safetyconcerned maneuvers of hard brake and cut-in. The simulation results show that the proposed model is collision free in the full-speed-range operation with leader accelerations within -1 to 1 m/s2 and in approaching and cut-out scenarios. Those results indicate that the proposed ACC-CACC car-following model can produce realistic vehicle response without causing vehicle collisions in regular scenarios for vehicle string operations.

Aggressive 3-D collision avoidance for high-speed navigation

Abstract: Autonomous robot navigation through unknown, cluttered environments at high-speeds is still an open problem. Quadrotor platforms with this capability have only begun to emerge with the advancements in light-weight, small form factor sensing and computing. Many of the existing platforms, however, require excessive computation time to perform collision avoidance, which ultimately limits the vehicle's top speed. This work presents an efficient perception and planning approach that significantly reduces the computation time by using instantaneous perception data for collision avoidance. Minimum-time, state and input constrained motion primitives are generated by sampling terminal states until a collision-free path is found. The worst case performance of the Triple Integrator Planner (TIP) is nearly an order of magnitude faster than the state-of-the-art. Experimental results demonstrate the algorithm's ability to plan and execute aggressive collision avoidance maneuvers in highly cluttered environments.

Collision Avoidance Strategies for Unmanned Aerial Vehicles in Formation Flight

Abstract: Collision avoidance strategies for multiple unmanned aerial vehicles (UAVs) based on geometry are investigated in this study. The proposed strategies allow a group of UAVs to avoid obstacles and separate if necessary through a simple algorithm with low computation by expanding the collision-cone approach to formation of UAVs. The geometric approach uses line-of-sight vectors and relative velocity vectors where dynamic constraints are included in the formation. Each UAV can determine which plane and direction are available for collision avoidance. An analysis is performed to define an envelope for collision avoidance, where angular rate limits and obstacle detection range limits are considered. Based on the collision avoidance envelope, each UAV in a formation determines whether the formation can be maintained or not while avoiding obstacles. Numerical simulations are performed to demonstrate the performance of the proposed strategies.

Modified social force model based on information transmission toward crowd evacuation simulation

Abstract: In this paper, the information transmission mechanism is introduced into the social force model to simulate pedestrian behavior in an emergency, especially when most pedestrians are unfamiliar with the evacuation environment. This modified model includes a collision avoidance strategy and an information transmission model that considers information loss. The former is used to avoid collision among pedestrians in a simulation, whereas the latter mainly describes how pedestrians obtain and choose directions appropriate to them. Simulation results show that pedestrians can obtain the correct moving direction through information transmission mechanism and that the modified model can simulate actual pedestrian behavior during an emergency evacuation. Moreover, we have drawn four conclusions to improve evacuation based on the simulation results; and these conclusions greatly contribute in optimizing a number of efficient emergency evacuation schemes for large public places.

Parallel autonomy in automated vehicles: Safe motion generation with minimal intervention

Abstract: Current state-of-the-art vehicle safety systems, such as assistive braking or automatic lane following, are still only able to help in relatively simple driving situations. We introduce a Parallel Autonomy shared-control framework that produces safe trajectories based on human inputs even in much more complex driving scenarios, such as those commonly encountered in an urban setting. We minimize the deviation from the human inputs while ensuring safety via a set of collision avoidance constraints. We develop a receding horizon planner formulated as a Non-linear Model Predictive Control (NMPC) including analytic descriptions of road boundaries, and the configurations and future uncertainties of other traffic participants, and directly supplying them to the optimizer without linearization. The NMPC operates over both steering and acceleration simultaneously. Furthermore, the proposed receding horizon planner also applies to fully autonomous vehicles. We validate the proposed approach through simulations in a wide variety of complex driving scenarios such as left-turns across traffic, passing on busy streets, and under dynamic constraints in sharp turns on a race track.

External Wrench Estimation, Collision Detection, and Reflex Reaction for Flying Robots

Abstract: Flying in unknown environments may lead to unforeseen collisions, which may cause serious damage to the robot and/or its environment. In this context, fast and robust collision detection combined with safe reaction is, therefore, essential and may be achieved using external wrench information. Also, deliberate physical interaction requires a control loop designed for such a purpose and may require knowledge of the contact wrench. In principle, the external wrench may be measured or estimated. Whereas measurement poses large demands on sensor equipment, additional weight, and overall system robustness, in this paper we present a novel model-based method for external wrench estimation in flying robots. The algorithm is based on the onboard inertial measurement unit and the robot's dynamics model only. We design admittance and impedance controllers that use this estimate for sensitive and robust physical interaction. Furthermore, the performance of several collision detection and reaction schemes is investigated in order to ensure collision safety. The identified collision location and associated normal vector located on the robot's convex hull may then be used for sensorless tactile sensing. Finally, a low-level collision reflex layer is provided for flying robots when obstacle avoidance fails, also under wind influence. Our experimental and simulation results show evidence that the methodologies are easily implemented and effective in practice.

A Novel Null-Space-Based UAV Trajectory Tracking Controller With Collision Avoidance

Abstract: A new trajectory tracking controller with collision avoidance is proposed in this paper for unmanned aerial vehicle (UAV) navigation. A positive potential function is designed to take into account the movement of obstacles. Thus, the controller with potential function guarantees that the UAV moves through areas of potentials close to zero to ensure safe navigation in dynamic and unknown environments. Such a controller was designed with hierarchical objectives using a behavioral-based approach. A null-space-based controller is adopted, whose main objective is to ensure that the collision avoidance is achieved, whereas other objectives are projected onto the null space. Collision avoidance and trajectory tracking controllers generate reference velocities sent to a dynamic compensator to guarantee the tracking of such velocities, thus characterizing a cascade controller. Stability of the entire closed-loop nonlinear system is demonstrated through Lyapunov's theory. A low-cost indoor framework with just one RGB-D sensor, which is a combination of a RGB (red-green-blue) camera with a depth sensor based on infrared light was used to estimate the positions of the UAV and obstacles. Simulation and experiments are run using a Parrot AR.Drone quadrotor and considering a person as a dynamic obstacle for an AR.Drone quadrotor, and some of their results are reported to validate the proposed controller.

Safe Certificate-Based Maneuvers for Teams of Quadrotors Using Differential Flatness

Abstract: Safety Barrier Certificates that ensure collision-free maneuvers for teams of differential flatness-based quadrotors are presented in this paper. Synthesized with control barrier functions, the certificates are used to modify the nominal trajectory in a minimally invasive way to avoid collisions. The proposed collision avoidance strategy complements existing flight control and planning algorithms by providing trajectory modifications with provable safety guarantees. The effectiveness of this strategy is supported both by the theoretical results and experimental validation on a team of five quadrotors.

Constrained iterative LQR for on-road autonomous driving motion planning

Abstract: There exist a lot of challenges in trajectory planning for autonomous driving: 1) Needs of both spatial and temporal planning for highly dynamic environments; 2) Nonlinear vehicle models and non-convex collision avoidance constraints. 3) High computational efficiency for real-time implementation. Iterative Linear Quadratic Regulator (ILQR) is an algorithm which solves predictive optimal control problem with nonlinear system very efficiently. However, it can not deal with constraints. In this paper, the Constrained Iterative LQR (CILQR) is proposed to handle the constraints in ILQR. Then an on road driving problem is formulated. Simulation case studies show the capability of the CILQR algorithm to solve the on road driving motion planning problem.

Smart vehicle monitoring and assistance using cloud computing in vehicular Ad Hoc networks

Abstract: The increasing number of on road vehicles has become a major cause for congestion, accidents and pollution. Intelligent Transportation Systems (ITS) might be the key to achieve solutions that help in reducing these problems significantly. The connected vehicular networks stream is a rapidly growing field for research and development of various real-time applications. In this paper, novel techniques have been proposed to serve the speed based lane changing, collision avoidance and time of arrival (TOA) based localization in Vehicular Ad Hoc Networks (VANETs). As GPS requires clear line-of-sight for accurate services of positioning and localization applications, we designed a Time of Arrival (ToA) based algorithm for areas where strong GPS signals are unavailable. Collision avoidance using automatic braking and camera-based surveillance are a few other applications that we addressed. The feasibility and the viability of the algorithms were demonstrated through simulations in Simulation of Urban Mobility (SUMO) and Network Simulator-2 (NS-2). We prototyped a working hardware and tested it on actual vehicles to assess the effectiveness of the proposed system. We designed a mobile app interface for the on-board unit for smart, efficient and remote traffic monitoring. The integrated VANET Cloud Computing architecture acts as the platform for the proposed applications.

Path planning and integrated collision avoidance for autonomous vehicles

Abstract: This paper discusses some of the current state-of-the-art and remaining challenges in research on path planning and vehicle control of autonomous vehicles. Reliable path planning is fundamental for the proper operation of an autonomous vehicle. Typically, the path planner relies on an incomplete model of the surroundings to generate a reference trajectory, used as input to a vehicle controller that tracks this reference trajectory. Depending on how much complexity is put into the path-planning block, the path planning and vehicle-control blocks can be viewed as independent of each other, connected to each other, or merged into one block. There are several types of path-planning techniques developed over the last decades, each with its own set of benefits and drawbacks. We review different techniques for path planning and trajectory tracking, and give examples of its use in relation to autonomous vehicles. We report on our own recent findings and give an outlook on potential research directions.

Collision and Deadlock Avoidance in Multirobot Systems: A Distributed Approach

Abstract: Collision avoidance is a critical problem in motion planning and control of multirobot systems. Moreover, it may induce deadlocks during the procedure to avoid collisions. In this paper, we study the motion control of multirobot systems where each robot has its own predetermined and closed path to execute persistent motion. We propose a real-time and distributed algorithm for both collision and deadlock avoidance by repeatedly stopping and resuming robots. The motion of each robot is first modeled as a labeled transition system, and then controlled by a distributed algorithm to avoid collisions and deadlocks. Each robot can execute the algorithm autonomously and real-timely by checking whether its succeeding state is occupied and whether the one-step move can cause deadlocks. Performance analysis of the proposed algorithm is also conducted. The conclusion is that the algorithm is not only practically operative but also maximally permissive. A set of simulations for a system with four robots are carried out in MATLAB. The results also validate the effectiveness of our algorithm.

System And Method Of Dynamically Controlling Parameters For Processing Sensor Output Data

Abstract: Various embodiments include dynamically controlling one or more parameters for obtaining and/or processing sensor data received from a sensor on a vehicle based on the speed of the vehicle. In some embodiments, parameters for obtaining and/or processing sensor data may be individually tuned (e.g., decreased, increased, or maintained) by leveraging differences in the level of quality, accuracy, confidence and/or other criteria in sensor data associated with particular missions/tasks performed using the sensor data. For example, the sensor data resolution required for collision avoidance may be less than the sensor data resolution required for inspection tasks, while the update rate required for inspection tasks may be less than the update rate required for collision avoidance. Parameters for obtaining and/or processing sensor data may be individually tuned based on the speed of the vehicle and/or the task or mission to improve consumption of power and/or other resources.

Distributed Stochastic Search Algorithm for Multi-ship Encounter Situations

Abstract: Ship collision avoidance involves helping ships find routes that will best enable them to avoid a collision. When more than two ships encounter each other, the procedure becomes more complex since a slight change in course by one ship might affect the future decisions of the other ships. Two distributed algorithms have been developed in response to this problem: Distributed Local Search Algorithm (DLSA) and Distributed Tabu Search Algorithm (DTSA). Their common drawback is that it takes a relatively large number of messages for the ships to coordinate their actions. This could be fatal, especially in cases of emergency, where quick decisions should be made. In this paper, we introduce Distributed Stochastic Search Algorithm (DSSA), which allows each ship to change her intention in a stochastic manner immediately after receiving all of the intentions from the target ships. We also suggest a new cost function that considers both safety and efficiency in these distributed algorithms. We empirically show that DSSA requires many fewer messages for the benchmarks with four and 12 ships, and works properly for real data from the Automatic Identification System (AIS) in the Strait of Dover.