Research and application of obstacle avoidance algorithm for hydropower station inspection robot based on spatio-temporal networks

Navigation and path planning are among the central problems in the development of mobile and autonomous robots. Research in this field has been conducted for decades, and several methodologies have been proposed to solve these problems. In the field, these approaches are divided into classical or deterministic and non-deterministic or heuristic methods. The article provides a brief overview of typical representatives of both classes, as well as an extended review of methods based on artificial potential fields. Important characteristics of obstacle detection and avoidance algorithms include convergence, computation time, and memory requirements in the system. The need for convergence arises from the requirement to achieve a stable or desired state of the system. This time varies depending on the chosen algorithm, the nature of the task, and the initial conditions. The main goal is to reduce convergence time, i.e., to reach the desired state as quickly as possible. Computation time and memory requirements are important because the robot must respond to the working environment and changes in it in real-time, and autonomous robots usually have quite limited hardware resources. Therefore, these are also important characteristics when selecting a method for a specific task and robot. The modification of the classical artificial potential field method using the Gaussian function to describe repulsive forces is an example of optimizing the method for systems with constrained resources. As of the writing of the article, unmanned aerial vehicles with limited resources are beginning to be widely used, making such optimizations practically valuable. Among the considered methods, heuristic ones are relatively new and are increasingly finding practical application. Research at the time of writing focuses on optimizing existing algorithms and hybridization to improve efficiency. An example of such hybridization is the artificial potential field method using fuzzy logic. This combines the classical artificial potential field method with a heuristic approach—fuzzy logic. This leads to some complexity in the method but solves typical problems of the classical algorithm, such as local minima, and increases the optimality and smoothness of the path. Most of obstacle detection and avoidance algorithms are working with only one type of sensor, such as ultrasonic distance sensors, LIDAR, or cameras. Each sensor technology and corresponding algorithms have their advantages and disadvantages. A promising approach is to use several types of sensors and algorithms, combining the results of different algorithms to achieve a more optimal final result, so called sensor fusion. However, it should be noted that this approach will require more sophisticated hardware. As robots increasingly become part of everyday life, it is quite possible that they will start working in collaboratively and interacting to solve assigned tasks. The development of collaborative methods for obstacle avoidance and interaction between robots in a single working environment is also a promising research direction. In summary, the gradual robotization of many processes in everyday life or production generates a high demand for research in the field of mobile robotics in general and methods for obstacle detection, avoidance and path planning in particular. Key words: robotics, obstacle avoidance, path planning, artificial potential field, autonomous robots, mobile robots.

An obstacle avoidance and path planning algorithm for a multi-joint manipulator in a space robot is presented in this paper. In this paper, the end-effector of the manipulator is used to capture some special target in a space environment with obstacles. To ensure the safety of the operation, a collision-free path from the initial position to the target position is essential. Therefore, an obstacle avoidance and path planning algorithm based on the Rapidly-Exploring Random Tree (RRT) algorithm and the Forward and Backward Reaching Inverse Kinematics (FABRIK) algorithm is presented in this paper. First, a path planning algorithm based on the Rapidly-Exploring Random Tree (RRT) algorithm is designed for the multi-joint manipulator. Further, a method to generate a random point by artificial guidance is introduced for a higher searching speed. The RRT algorithm can effectively explore the entire workspace and find a feasible path without collision for the end-effector. To calculate the positions of each joint, the Forward and Backward Reaching Inverse Kinematics (FABRIK) algorithm is introduced and improved for the problem of inverse kinematics. The FABRIK algorithm avoids the use of rotational angles or matrices, and instead finds each joint position by locating a point on a line, and thus, it has a low computational cost. Therefore, the improved obstacle avoidance and path planning algorithm can quickly plan a feasible path for the multi-joint manipulator in a space environment with obstacles. A numerical simulation is carried out to analyze the proposed obstacle avoidance and path planning method. It is observed that the method finds a feasible path without collision for the multi-joint manipulator with a low computational cost. These results validated the effectiveness of the proposed method for path planning to avoid the obstacles.

Due to the complexity of the ocean environment, an autonomous underwater vehicle (AUV) is disturbed by obstacles when performing tasks. Therefore, the research on underwater obstacle detection and avoidance is particularly important. Based on the images collected by a forward-looking sonar on an AUV, this article proposes an obstacle detection and avoidance algorithm. First, a deep learning-based obstacle candidate area detection algorithm is developed. This algorithm uses the You Only Look Once (YOLO) v3 network to determine obstacle candidate areas in a sonar image. Then, in the determined obstacle candidate areas, the obstacle detection algorithm based on the improved threshold segmentation algorithm is used to detect obstacles accurately. Finally, using the obstacle detection results obtained from the sonar images, an obstacle avoidance algorithm based on deep reinforcement learning (DRL) is developed to plan a reasonable obstacle avoidance path of an AUV. Experimental results show that the proposed algorithms improve obstacle detection accuracy and processing speed of sonar images. At the same time, the proposed algorithms ensure AUV navigation safety in a complex obstacle environment.

It is only with time that the efficiency or the effectiveness of the algorithms for obstacle avoidance gets better, and experiences of any kind can be inferred for the betterment of the knowledge on this domain. For a mobile robot navigating its way from starting point to an ending point while traversing through deterrents, needs to divide the problem into sub problems. It fundamentally involves, sensory data interpretation, choosing apt algorithm based on the objective function, and configuring the mobile robot accordingly to attain desired output. In this paper, few essential classifications for obstacle avoidance and robot navigation algorithms are discussed. Importance of the hardware aspect of the robot is undeniable. A set of algorithms were classified into 2 main classifications which are further divided into sub classifications in an arranged and concise manner. This information can be used to develop a suitable model for the given problem. Understanding fundamental ideas or strategies would allow in developing a novel extended strategy, although few specialized strategies are saturated, nevertheless can still be used as valuable alternatives. These alternatives may involve algorithms which have paramount potential, per se which are interestingly similar to the functioning of a brain, nature-inspired, etc.

Aiming at the problem that obstacle avoidance algorithms do not distinguish the types of obstacles in the process of obstacle recognition and avoidance, this paper designs a robot obstacle avoidance system including knowledge base by combining prior knowledge with obstacle avoidance algorithm. The knowledge in the knowledge base can be defined by human as prior knowledge to classify obstacles. This makes it possible for the robot to identify obstacles according to the vision sensor in the obstacle avoidance process, and can choose different obstacle avoidance strategies for different obstacles according to the existing knowledge in the knowledge base. The knowledge base takes knowrob as the platform, and the experimental results show that the obstacle avoidance system can effectively avoid obstacles.

Obstacle avoidance for orchard vehicle trinocular vision system based on coupling of geometric constraint and virtual force field method

Multi-vehicle formation control and obstacle avoidance using negative-imaginary systems theory

Deep spatiotemporal fusion network for vision-based robotic inspection of structures

Research is underway at the NASA Glenn Researc h Center to develop and demonstrate the core technologies required to enable new and revolutionary approaches to engine diagnostics. One such effort investigates the use of multi -agent robots for autonomous search and inspection of propulsion systems. In t his envisioned application, on -wing engine inspections will be performed by groups of miniature mobile inspection devices that will traverse the interior surfaces of engine components in a coordinated, comprehensive search and inspection for damage. The cu rrent effort consists of several parallel activities that include: investigation into a range of algorithms for cooperative search, coverage completeness and obstacle avoidance; development of 3 -D graphical simulation software that will serve as a virtual test -bed to facilitate the testing and validation of the control algorithms; and development of demonstration robots that will allow the integration of control algorithms onto hardware. These activities will culminate in a proof -of -concept demonstration in which inspection robots will work together in a coordinated fashion to search for damage targets in a hardware test -bed environment. This demonstration will validate the feasibility of using multi -agent robotics to perform cooperative inspections in app lications such as on -wing in situ turbine engine maintenance .

The availability of powerful eye-safe laser sources and the recent advancements in electro-optical and mechanical beam-steering components have allowed laser-based Light Detection and Ranging (LIDAR) to become a promising technology for obstacle warning and avoidance in a variety of manned and unmanned aircraft applications. LIDAR outstanding angular resolution and accuracy characteristics are coupled to its good detection performance in a wide range of incidence angles and weather conditions, providing an ideal obstacle avoidance solution, which is especially attractive in low-level flying platforms such as helicopters and small-to-medium size Unmanned Aircraft (UA). The Laser Obstacle Avoidance Marconi (LOAM) system is one of such systems, which was jointly developed and tested by SELEX-ES and the Italian Air Force Research and Flight Test Centre. The system was originally conceived for military rotorcraft platforms and, in this paper, we briefly review the previous work and discuss in more details some of the key development activities required for integration of LOAM on UA platforms. The main hardware and software design features of this LOAM variant are presented, including a brief description of the system interfaces and sensor characteristics, together with the system performance models and data processing algorithms for obstacle detection, classification and avoidance. In particular, the paper focuses on the algorithm proposed for optimal avoidance trajectory generation in UA applications.

<p>The availability of powerful eye-safe laser sources and the recent advancements in electro-optical and mechanical beam-steering components have allowed laser-based Light Detection and Ranging (LIDAR) to become a promising technology for obstacle warning and avoidance in a variety of manned and unmanned aircraft applications. LIDAR outstanding angular resolution and accuracy characteristics are coupled to its good detection performance in a wide range of incidence angles and weather conditions, providing an ideal obstacle avoidance solution, which is especially attractive in low-level flying platforms such as helicopters and small-to-medium size Unmanned Aircraft (UA). The Laser Obstacle Avoidance Marconi (LOAM) system is one of such systems, which was jointly developed and tested by SELEX-ES and the Italian Air Force Research and Flight Test Centre. The system was originally conceived for military rotorcraft platforms and, in this paper, we briefly review the previous work and discuss in more details some of the key development activities required for integration of LOAM on UA platforms. The main hardware and software design features of this LOAM variant are presented, including a brief description of the system interfaces and sensor characteristics, together with the system performance models and data processing algorithms for obstacle detection, classification and avoidance. In particular, the paper focuses on the algorithm proposed for optimal avoidance trajectory generation in UA applications.</p>

The paper deals with kinematic control algorithms for on-line obstacle avoidance which allow a kinematically redundant manipulator to move in an unstructured environment without colliding with obstacles. The presented approach is based on the redundancy resolution at the velocity level. The primary task is determined by the end-effector trajectories and for obstacle avoidance the internal motion of the manipulator is used. The obstacle avoiding motion is defined in one-dimensional operational space and, hence, the system has less singularities making implementation easier. Instead of the exact pseudoinverse solution we propose an approximate one which is computationally more efficient and allows us to consider many simultaneously active obstacles without any problems. The fast cycle times of the numerical implementation enable use of the algorithm in real-time control. For illustration, some simulation results of a highly redundant planar manipulator moving in an unstructured and time-varying environment and experimental results of a four link planar manipulator are given.

A Decentralized Control Algorithm for Swarm Behavior and Obstacle Avoidance in Unknown Environments

In this paper, the development and demonstration of a modified Voronoi algorithm for unmanned aerial vehicle (UAV) path planning and obstacle avoidance is presented. This algorithm is intended to produce a flyable collision-free path through a series of obstacles/threats represented by cylindrical risk zones. A model of the West Virginia University (WVU) YF-22 research aircraft implemented within the WVU UAV simulation environment is used to demonstrate the functionality of the proposed algorithm and its effectiveness in the presence of several risk zones configurations. As compared to traditional methods, the approach is more general, coping with obstacles and risk zones of variable intensity and providing additional flexibility and better coverage of the possible solution space.

The robustness of obstacle avoidance algorithm is one of the important factors to successful applications of mobile robot systems. The sonar ring is used widely for autonomous mobile robot obstacle avoidance. This paper first analyzes the robustness of the existing obstacle avoidance algorithms based on sonar ring, indicates that the certainty grid method for obstacle representation is helpful to the robustness improvement of obstacle avoidance algorithms, but its effect is limited, it also has many disadvantages. By the simulation of two typical obstacle avoidance algorithms, the damage of interfered sonar data is revealed. Then the kinematics model of obstacle avoidance is built, Kalman filter which can restrain divergence is designed for interfered sonar data. Sonar data is used by obstacle avoidance algorithm after filtering. By the simulation contrast of the two obstacle avoidance algorithms, the effect of the Kalman filter for robustness improvement of obstacle avoidance algorithms is testified. Finally, the effect of the Kalman filter for eliminating noises in sonar data and for robustness improvement of obstacle avoidance algorithms is verified by experiments in two different situation.