Obstacle Avoidance Performance Research Articles

Event‐triggered obstacle avoidance control for autonomous surface vehicles with actuator faults

AbstractThe article addresses the event‐triggered obstacle avoidance control problem for autonomous surface vehicles subject to actuator faults. In order to tackle the challenges presented by unknown actuator faults, internal model uncertainties, and external disturbances, we propose a fault‐tolerant control scheme that relies on an extended state observer to estimate the unknown parameters. In addition, considering the occurrence of actuator faults, a novel event‐triggered mechanism is developed to reduce wear of actuator and ensures the performance of system. On this basis, we employ backstepping technique and an improved artificial potential function methods to develop an event‐triggered obstacle avoidance control scheme with the capability of fault tolerance. This proposed control strategy ensures the uniform ultimate boundedness of tracking errors. In contrast to existing results, the presented control strategy simultaneously holds the performance of obstacle avoidance and fault tolerance and reduce the update frequency of actuators. The effectiveness of the provided control scheme has been confirmed through simulation.

A Gaussian-Process-Based Model Predictive Control Approach for Trajectory Tracking and Obstacle Avoidance in Autonomous Underwater Vehicles

To achieve the efficient and precise control of autonomous underwater vehicles (AUVs) in dynamic ocean environments, this paper proposes an innovative Gaussian-Process-based Model Predictive Control (GP-MPC) method. This method combines the advantages of Gaussian process regression in modeling uncertainties in nonlinear systems, and MPC’s constraint optimization and real-time control abilities. To validate the effectiveness of the proposed GP-MPC method, its performance is first evaluated for trajectory tracking control tasks through numerical simulations based on a 6-degrees-of-freedom, fully actuated, AUV dynamics model. Subsequently, for 3D scenarios involving static and dynamic obstacles, an AUV horizontal plane decoupled motion model is constructed to verify the method’s obstacle avoidance capability. Extensive simulation studies demonstrate that the proposed GP-MPC method can effectively manage the nonlinear motion constraints faced by AUVs, significantly enhancing their intelligent obstacle avoidance performance in complex dynamic environments. By effectively handling model uncertainties and satisfying motion constraints, the GP-MPC method provides an innovative and efficient solution for the design of AUV control systems, substantially improving the control performance of AUVs.

Hierarchical path planner combining probabilistic roadmap and deep deterministic policy gradient for unmanned ground vehicles with non-holonomic constraints

Path planning plays a vital role in autonomous driving as it needs to guide the vehicle to achieve the target position without collision while satisfying the vehicle's non-holonomic constraints. A good path planner should be able to cope with sophisticated environments and try to make the planned path as short as possible. Traditional path planning methods when being applied on four-wheel vehicles may suffer from one of the following disadvantages including conflicting with non-holonomic constraints, far from close-optimal length solution, or easy to get stuck in specific environments. In this paper, a hierarchical path planning method for a four-wheel vehicle combining probabilistic roadmap (PRM) and deep deterministic policy gradient (DDPG) is proposed to address existing problems. In the upper level, PRM is used to generate a guidance path quickly, which doesn't consider the vehicle's non-holonomic constraints but gives the sketch of the path that helps the vehicle to jump out of some dilemmas in specific environments while maintaining a relatively short path distance. In the lower level, reinforcement learning, specifically DDPG, is used to optimize the guidance path generated by PRM considering multiple factors including non-holonomic constraints and obstacle avoidance. The proposed method is validated in four different environments including a simple map, an easy-to-stuck specific map, a maze-like complex map and an office-like complex map. Results show that the proposed method could generate a smooth path satisfying non-holonomic constraints while demonstrating desirable obstacle avoidance performance. The planned path of the proposed method is superior to that of both traditional and state-of-the-art methods in terms of path length and smoothness, which could be 1.3–12.9 % shorter and 2.6–52.1 % smoother in different environments.

Multimodal audio-visual robot fusing 3D CNN and CRNN for player behavior recognition and prediction in basketball matches.

Intelligent robots play a crucial role in enhancing efficiency, reducing costs, and improving safety in the logistics industry. However, traditional path planning methods often struggle to adapt to dynamic environments, leading to issues such as collisions and conflicts. This study aims to address the challenges of path planning and control for logistics robots in complex environments. The proposed method integrates information from different perception modalities to achieve more accurate path planning and obstacle avoidance control, thereby enhancing the autonomy and reliability of logistics robots. Firstly, a 3D convolutional neural network (CNN) is employed to learn the feature representation of objects in the environment for object recognition. Next, long short-term memory (LSTM) is used to model spatio-temporal features and predict the behavior and trajectory of dynamic obstacles. This enables the robot to accurately predict the future position of obstacles in complex environments, reducing collision risks. Finally, the Dijkstra algorithm is applied for path planning and control decisions to ensure the robot selects the optimal path in various scenarios. Experimental results demonstrate the effectiveness of the proposed method in terms of path planning accuracy and obstacle avoidance performance. The method outperforms traditional approaches, showing significant improvements in both aspects. The intelligent path planning and control scheme presented in this paper enhances the practicality of logistics robots in complex environments, thereby promoting efficiency and safety in the logistics industry.

Deep Reinforcement Learning for Autonomous Driving in Amazon Web Services DeepRacer

The development of autonomous driving models through reinforcement learning has gained significant traction. However, developing obstacle avoidance systems remains a challenge. Specifically, optimising path completion times while navigating obstacles is an underexplored research area. Amazon Web Services (AWS) DeepRacer emerges as a powerful infrastructure for engineering and analysing autonomous models, providing a robust foundation for addressing these complexities. This research investigates the feasibility of training end-to-end self-driving models focused on obstacle avoidance using reinforcement learning on the AWS DeepRacer autonomous race car platform. A comprehensive literature review of autonomous driving methodologies and machine learning model architectures is conducted, with a particular focus on object avoidance, followed by hands-on experimentation and the analysis of training data. Furthermore, the impact of sensor choice, reward function, action spaces, and training time on the autonomous obstacle avoidance task are compared. The results of the best configuration experiment demonstrate a significant improvement in obstacle avoidance performance compared to the baseline configuration, with a 95.8% decrease in collision rate, while taking about 79% less time to complete the trial circuit.

Path planning of unmanned vehicles based on adaptive particle swarm optimization algorithm

Path planning technology is the basis of autonomous driving of unmanned vehicles. However, there are some problems in the traditional path planning technology. For example, high-quality global paths can't be generated quickly; Lacking of security verification ability; The performance of dynamic obstacle avoidance is poor. Therefore, this paper proposes a path planning method of unmanned vehicles based on adaptive particle swarm optimization algorithm (APSO). Firstly, a map simplification strategy (MSS) is proposed. The grid map is preprocessed by map simplification strategy to reduce the search space and time of path planning algorithm; Secondly, an APSO algorithm is proposed. The algorithm coordinates the search of particles through three adaptive factors and Levy flight strategy. Then, a security checking strategy is proposed. Security checking strategy can be used to verify the safety of global path; Finally, a dynamic obstacle avoidance strategy based on behavior is proposed. Vehicles can independently analyze the types of collision and adopt corresponding obstacle avoidance strategies. The simulation results show that MSS-APSO algorithm and APSO algorithm surpass original algorithms and comparison algorithms; MSS-APSO algorithm has strong applicability in real map environment; The obstacle avoidance strategy has great obstacle avoidance ability and real-time performance; The map simplification strategy can improve iterations of the algorithm and quality of the global path.

Obstacle avoidance decision and trajectory tracking control of intelligent vehicle considering surrounding vehicles

To enhance the obstacle avoidance performance and trajectory tracking control stability of vehicles in emergency scenarios, and to solve the problem that the existing decision-making methods based on the minimum safe distance model cannot effectively make decisions in some scenarios, a novel lane-changing obstacle avoidance decision-making and control method is proposed. Initially, when emergency braking cannot avoid obstacles safely, the lane-changing trajectory is planned by quintic polynomial, and the minimum distance and collision detection (MDCD) algorithm is then employed to ascertain whether the lane-changing trajectory meets the safety conditions. Subsequently, a Linear Time-Varying Model Predictive Control (LTV MPC) trajectory tracking controller is designed based on the 7-DOF vehicle dynamics model. Finally, the effectiveness of the proposed method is verified through two representative scenarios. The simulation results indicate that the MDCD algorithm can ensure the safety and effectiveness of obstacle avoidance decisions in all emergency obstacle avoidance scenarios, and when tracking the emergency lane-changing trajectory in different typical scenarios, compared with the LTV MPC controller based on the single-track vehicle dynamics model (STVDM), the maximum tracking error can be reduced by 50.7% and 60.1%, respectively, while ensuring near-equivalent real-time performance operation.

Obstacle avoidance approaches of autonomous mobile robots in harsh environment

Obstacle avoidance has been a hotspot in the field of robotics. Good performance of obstacle avoidance is key to the success of robots, especially in harsh environment where the external environmental factors pose challenges to the robots. Currently, numerous obstacle avoidance methods are being developed in an attempt to optimize the decision of the autonomous mobile robots in harsh environment. To enhance the reliability of obstacle avoidance methods, it is essential to have a clear picture of the advances and breakthroughs in this field and see which research area needs to be improved. This paper focuses on the current obstacle avoidance methods or path-planning algorithms in the underwater environment and the space environment, with description of their works and discussion of their advantages and weakness. Current problems of these methods include lack of reliable field tests, inadequate assumption of obstacles when designing the algorithm and limited capability in handling special cases that no possible path is feasible. This paper intends to provide researchers with a generalized review of obstacle avoidance methods in harsh environment and a proposal of the potential research directions in this field in the future. It is recommended that researchers to design more complex scenarios, improve the performance of robots to avoid being trapped in special situations, and consider the adaptability of robots in different environment.

Global Time-Varying Path Planning Method Based on Tunable Bezier Curves

In this paper, a novel global time-varying path planning (GTVP) method is proposed. In the method, real-time paths can be generated based on tunable Bezier curves, which can realize obstacle avoidance of manipulators. First, finite feature points are extracted to represent the obstacle information according to the shape information and position information of the obstacle. Then, the feature points of the obstacle are converted into the feature points of the curve, according to the scale coefficient and the center point of amplification. Furthermore, a Bezier curve representing the motion path at this moment is generated to realize real-time adjustment of the path. In addition, the 5-degree Bezier curve planning method consider the start direction and the end direction is used in the path planning to avoid the situation of abrupt change with oscillation of the trajectory. Finally, the GTVP method is applied to multi-obstacle environment to realize global time-varying dynamic path planning. Through theoretical derivation and simulation, it can be proved that the path planned by the GTVP method can meet the performance requirements of global regulation, real-time change and multi-obstacle avoidance simultaneously.

A dynamic obstacle avoidance method for collaborative robots based on trajectory optimization

Background Collision detection is crucial in the design of robot planning algorithms. Efficient distance sensors can provide high-resolution environmental collision information to the robot's planning algorithm. However, this also leads to the robot obstacle avoidance performance being limited by the performance of the sensors. Therefore, it becomes a challenge to achieve efficient obstacle avoidance with low-resolution environmental information. Methods First, we use a self-developed capacitive array non-contact distance sensing flexible surface for sensing the proximity of colliding objects. Second, we designed an optimization-based dynamic obstacle avoidance planning algorithm, using only the minimum separation distance and penetration direction as obstacle avoidance information, and referring to the idea of stochastic gradient descent, using real-time collision avoidance information to do single-step optimization adjustment. Results We conducted the dynamic obstacle avoidance test experiment by connecting the electronic skin to the semi-physical prototype and the full physical prototype. The experiments show that efficient dynamic obstacle avoidance can be realized under the maximum effective range of only 5~7cm, and it has strong flexibility to avoid different shapes of dynamic obstacles in a non-contact manner, and finally arrive at the target position. Conclusions In this paper, an online obstacle avoidance planning algorithm designed based on an optimization method that is not limited to the shape of obstacles is proposed, and the effectiveness of the algorithm is verified by physical experiments in combination with a self-developed flexible distance sensing surface. It is of great significance for the safe operation of human-robot interaction in collaborative robots.

A Versatile Approach to Polygonal Object Avoidance in Indoor Environments with Hardware Schemes Using an FPGA-Based Multi-Robot

Service robots perform versatile functions in indoor environments. This study focuses on obstacle avoidance using flock-type indoor-based multi-robots. Each robot was developed with rendezvous behavior and distributed intelligence to perform obstacle avoidance. The hardware scheme-based obstacle-avoidance algorithm was developed using a bio-inspired flock approach, which was developed with three stages. Initially, the algorithm estimates polygonal obstacles and their orientations. The second stage involves performing avoidance at different orientations of obstacles using a heuristic based Bug2 algorithm. The final stage involves performing a flock rendezvous with distributed approaches and linear movements using a behavioral control mechanism. VLSI architectures were developed for multi-robot obstacle avoidance algorithms and were coded using Verilog HDL. The novel design of this article integrates the multi-robot’s obstacle approaches with behavioral control and hardware scheme-based partial reconfiguration (PR) flow. The experiments were validated using FPGA-based multi-robots.

Directly-trained Spiking Neural Networks for Deep Reinforcement Learning: Energy efficient implementation of event-based obstacle avoidance on a neuromorphic accelerator

Spiking Neural Networks (SNN) promise extremely low-power and low-latency inference on neuromorphic hardware. Recent studies demonstrate the competitive performance of SNNs compared with Artificial Neural Networks (ANN) in conventional classification tasks.In this work, we present an energy-efficient implementation of a Reinforcement Learning (RL) algorithm using SNNs to solve an obstacle avoidance task performed by an Unmanned Aerial Vehicle (UAV), taking a Dynamic Vision Sensor (DVS) as event-based input. We train the SNN directly, improving upon state-of-art implementations based on hybrid (not directly trained) SNNs. For this purpose, we devise an adaptation of the Spatio-Temporal Backpropagation algorithm (STBP) for RL. We then compare the SNN with a state-of-art Convolutional Neural Network (CNN) designed to solve the same task. To this aim, we train both networks by exploiting a photorealistic training pipeline based on AirSim. To achieve a realistic latency and throughput assessment for embedded deployment, we designed and trained three different embedded SNN versions to be executed on state-of-art neuromorphic hardware, targeting state-of-the-art.We compared SNN and CNN in terms of obstacle avoidance performance showing that the SNN algorithm achieves better results than the CNN with a factor of 6× less energy. We also characterize the different SNN hardware implementations in terms of energy and spiking activity.

A Motion Planning Method for Robots Based on DMPs and Modified Obstacle-Avoiding Algorithm

This paper addresses the motion planning of the manipulator in task space. To improve the overall trajectory performance, a special motion planning method based on DMPs (Dynamic Movement Primitives) and the modified obstacle-avoiding algorithm is proposed. The proposed method solves the problems of trajectory jitter and inability to avoid obstacles in some scenarios, which are faced by the scheme of steering angle. Besides, it improves the retention of teaching intentions, reduces the loss of free space, and helps the system adapt to the dynamic environment. At the theoretical level, the convergence of the target state has been proved using Lyapunov stability theory-based analysis. The availability of the proposed method is validated and analyzed by performing a series of numerical simulations and Baxter robot experiments. The results indicate that the proposed method can provide reliable solutions for motion planning. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —From the perspective of demonstration learning, this paper aims to elevate the motion planning of the manipulator in task space, especially in improving the obstacle avoidance performance. The existing DMPs-based motion planning algorithm uses the scheme of steering angle to achieve obstacle avoidance, and the obstacle is regarded as a mesh of points on the boundary. The obstacle-avoiding performance is limited. How to combine the obstacle-avoiding algorithm with the DMPs-based motion planning algorithm more effectively, so as to simultaneously achieve retaining teaching intentions as much as possible, still faces challenges. This paper designs a modified obstacle-avoiding algorithm, which solves the problems of trajectory jitter and inability to avoid obstacles in some circumstances, and improves the retention of teaching intentions. The modified obstacle-avoiding algorithm is also applied to the dynamic environment. The proposed method satisfies Lyapunov stability theory-based analysis, and the experiments on Baxter robot verify the feasibility.

Robot Path Planning Research Incorporating Improved A* Algorithm and DWA Algorithm

For the traditional A* algorithm has problems such as long paths, large number of nodes, and the demand for dynamic obstacle cannot be avoided in complex environment. A mobile robot dynamic path avoidance method will be improved to improve the A * algorithm and improve DWA algorithm Two map environments are used for simulation verification. First, the evaluation function and key node selection strategy are optimized for the A* algorithm, and redundant nodes are deleted; then the dynamic obstacle distance evaluation function is added to the DWA algorithm which for the purpose of the obstacle avoidance performance can be enhanced. The results about the improved A* algorithm reduces 12.20% and 58.33% in path length and number of turning points respectively compared with the traditional A* algorithm can be obviously grasped by the simulation experiment; by using the fusion algorithm whose purpose of using arcs instead of the straight lines is to turn more smoothly, and can be closest to the global optimum while avoiding dynamic obstacles to complete the search.

Improved path planning method for unmanned aerial vehicles based on artificial potential field

With the advancement of unmanned aerial vehicle (UAV) technology and its widespread use in many facets of manufacturing and daily life, the need for UAV mission automation is becoming more and more practical. In order to improve the automatic obstacle avoidance and path planning performance of UAVs, this essay proposes an optimized route planning algorithm based on the artificial potential field (APF) method, which have solved the typical issue of the APF. This method chooses to conduct a pre planning trajectory of the UAV based on a rapidly expanding random tree (RRT). The pre-planned path will be split into continuous particles, and then generating intermediate waypoints. A nearby waypoint offers a gravitational force to aid the UAV in escaping the local minimum when it enters it. At the same time, taking the distance from the UAV to the obstacle and the radius of influence of the obstacle itself into consideration, dynamically adjust the gravitational and repulsive coefficients, set up a non-gravitational zone around the obstacle, which is beneficial for the UAV to elude the obstacle. And set up a repulsive limited action zone to reduce unnecessary turns in the trajectory to achieve the effect of path optimization. Considering the difficulty of single UAV missions in most cases, this paper discusses cooperative flight path planning for multiple UAVs into consideration.

EMPMR berthing scheme: A novel event-triggered motion planning and motion replanning scheme for unmanned surface vessels

In this study, the issues related to approaching and berthing of unmanned surface vessels (USVs) are investigated. The automatic berthing task is divided into motion planning and motion replanning. First, a smooth path from the initial position to the berth is planned using A* algorithm and B-spline. Then, the segmented interpolation method is employed to let the USV velocities converge slowly to zero at the berth, and an initial berthing trajectory is obtained. Moreover, considering the kinetic constraints of the USV, the nonlinear model predictive control (NMPC) algorithm is employed in the offline motion planning and online motion replanning scheme. In this scheme, novel static, and COLREGs-based dynamic obstacle avoidance functions are proposed based on Gaussian functions. Furthermore, the event-triggered mechanism is adopted to improve the real-time performance of the NMPC algorithm. The proposed event-triggered motion planning and motion replanning (EMPMR) scheme and Han et al.’s scheme are tested in six scenarios. Simulation results show that the proposed EMPMR scheme can accomplish USV’s approaching and automatic berthing of USVs and has better obstacle avoidance and computational performance.

A FUZZY CONTROLLER FOR A MOBILE ROBOT WITH OBSTACLE AVOIDANCE

This research paper presents a mobile robot based on the MAMDANI fuzzy inference system. The mobile robot is equipped with three pairs of HC-SR04 ultrasonic sensors to detect the distance between the robot and obstacles. A fuzzy logic controller is built into the STM32F4 microcontroller platform to generate the actuation signal for the DC motor mounted on each robot wheel. To perform obstacle avoidance, a sensor capable of detecting the distance between the robot and the obstacle must be installed so that the controller can calculate and determine the appropriate control signal to send to the robot's actuator. In achieving this goal, the robot can perform movements at a constant speed or at a variable speed. When the distance between the robot and the obstacle is too far, the speed is too fast. Then, as the distance to the obstacle increases, the speed decreases. Such a control mechanism can be implemented with certain control methods installed in the controller. The data obtained from the ultrasonic sensors are expressed as linguistic variables that indicate the speed of movement of each wheel in fuzzy sets. A robot with a fuzzy logic controller based on a rule table is designed to achieve uninterrupted motion with obstacle avoidance capability. Keywords: STM32F4 microcontroller, Fuzzy logic, Fuzzy inference system, mobile robot, obstacle avoidance.

Application of AI and ML

Artificial inteligence(AI) and machine learning(ML) are arguably the most distruptive technologies in the modern world. The possible applications in robotics and automation are virtually limitless. Thus, this paper explores uses in two key areas: servo motors and the kinematics and dynamics models of robotics and drones. Reliable studies providing insights into these topics were used to develop the report. The study found that AI and ML enable precise control, obstacle avoidance, and adaptive performance in servo motors and motion planning, snesor fusion, fault detection, and system identification in kinematics and dynamics models. These findings confirm the vitality of AI and ML for the functionality of intelligent robots and drones used across different industries, such as construction and manufacturing.

Intelligent Mining Road Object Detection Based on Multiscale Feature Fusion in Multi-UAV Networks

In complex mining environments, driverless mining trucks are required to cooperate with multiple intelligent systems. They must perform obstacle avoidance based on factors such as the site road width, obstacle type, vehicle body movement state, and ground concavity-convexity. Targeting the open-pit mining area, this paper proposes an intelligent mining road object detection (IMOD) model developed using a 5G-multi-UAV and a deep learning approach. The IMOD model employs data sensors to monitor surface data in real time within a multisystem collaborative 5G network. The model transmits data to various intelligent systems and edge devices in real time, and the unmanned mining card constructs the driving area on the fly. The IMOD model utilizes a convolutional neural network to identify obstacles in front of driverless mining trucks in real time, optimizing multisystem collaborative control and driverless mining truck scheduling based on obstacle data. Multiple systems cooperate to maneuver around obstacles, including avoiding static obstacles, such as standing and lying dummies, empty oil drums, and vehicles; continuously avoiding multiple obstacles; and avoiding dynamic obstacles such as walking people and moving vehicles. For this study, we independently collected and constructed an obstacle image dataset specific to the mining area, and experimental tests and analyses reveal that the IMOD model maintains a smooth route and stable vehicle movement attitude, ensuring the safety of driverless mining trucks as well as of personnel and equipment in the mining area. The ablation and robustness experiments demonstrate that the IMOD model outperforms the unmodified YOLOv5 model, with an average improvement of approximately 9.4% across multiple performance measures. Additionally, compared with other algorithms, this model shows significant performance improvements.

Dynamic Obstacle Avoidance for USVs Using Cross-Domain Deep Reinforcement Learning and Neural Network Model Predictive Controller

This work presents a framework that allows Unmanned Surface Vehicles (USVs) to avoid dynamic obstacles through initial training on an Unmanned Ground Vehicle (UGV) and cross-domain retraining on a USV. This is achieved by integrating a Deep Reinforcement Learning (DRL) agent that generates high-level control commands and leveraging a neural network based model predictive controller (NN-MPC) to reach target waypoints and reject disturbances. A Deep Q Network (DQN) utilized in this framework is trained in a ground environment using a Turtlebot robot and retrained in a water environment using the BREAM USV in the Gazebo simulator to avoid dynamic obstacles. The network is then validated in both simulation and real-world tests. The cross-domain learning largely decreases the training time (28%) and increases the obstacle avoidance performance (70 more reward points) compared to pure water domain training. This methodology shows that it is possible to leverage the data-rich and accessible ground environments to train DRL agent in data-poor and difficult-to-access marine environments. This will allow rapid and iterative agent development without further training due to the change in environment or vehicle dynamics.