A deep reinforcement learning based method for real-time path planning and dynamic obstacle avoidance

A system for placing objects on a surface. The system may include a base, a robotic arm coupled, at an end thereof, to the base, an end effector coupled to the other end of the robotic arm. The end effector may be configured for releaseably coupling to an object to be placed on the surface. The system may further include one or more sensor units on a sensor frame. The one or more sensor units may be configured for sensing a two-dimensional profile data including at least two two-dimensional profiles together comprising at least three boundary portions of the object to be placed and at least three boundary portions of objects on the surface. At least two of the three boundary portions of the object to be placed may be from substantially non-parallel sides. At least two of the three boundary portions of the objects on the surface may be from substantially non-parallel sides. The system may further include a processor configured to determine at least three degrees of freedom of the object to be placed with respect to the sensor frame and six degrees of freedom of the sensor frame with respect to the objects on the surface in a three-dimensional space for determining a current pose of the object to be placed with respect to the objects on the surface based on the two-dimensional profile data. Further, the system may be configured to place the object based on differences between the current pose and a desired pose of the object to be placed determined from a model of objects on the surface in the three-dimensional space.

System for placing objects on a surface and method thereof

This paper presents a real-time method for whole-body trajectory optimization of mobile manipulators in simplified dynamic and unstructured environments. Current trajectory optimization methods typically use decoupling of the mobile base and the robotic arm, which reduces flexibility in motion, does not scale to unstructured environments, and does not consider the future evolution of the environment, which is crucial to avoid dynamic obstacles. Given a goal configuration, such as waypoints generated by a global path planner, we formulate a receding horizon trajectory optimization minimizing the distance-to-target while avoiding collisions with static and dynamic obstacles. The presented method unifies the control of a robotic arm and a non-holonomic base to allow coupled trajectory planning. For collision avoidance, we propose to compute three convex regions englobing the robot's major body parts (i.e., base, shoulder-link and wrist-link) and thus reducing and limiting the number of inequality constraints, regardless of the number of obstacles in the environment. Moreover, our approach incorporates predicted trajectory information to smoothly, and in advance, avoid dynamic obstacles. The presented results show that trajectory optimization for the coupled system can reduce the total execution time by 48% and that applying the convex region generation for individual links allows keeping the computational costs low, even for complex scenarios, enabling onboard implementation.

Coupled Mobile Manipulation via Trajectory Optimization with Free Space Decomposition

The invention discloses a vision based robot demonstration and learning method, and the method utilizes a demonstration tool and a vision sensor to perform learning of a demonstration task. A demonstrator firstly holds the demonstration tool in hands to demonstrate an operation tank, the vision sensor obtains image characteristics of the demonstration tool, and a demonstration track, in a demonstration process, of the demonstration tool is obtained according to inner parameters of the vision sensor, so that robot motion is controlled to obtain a tail-end motion track, in a demonstration process, of the robot is obtained; and Kalman filtering is finally performed on the robot tail-end motion track to obtain learning track of the robot, so that learning of the demonstration task is realized.By virtue of a simple vision tool, six-dimensional pose information of the demonstration tool is easily extracted, so that real-time property of demonstration and learning is good. The vision based robot demonstration and learning method reduces demonstration difficulty of an operator, so that the operator with no experience also can perform demonstration and teaching of the robot.

Vision based robot demonstration and learning method

The invention discloses a virtual velocity vector field-based robot three-dimensional collision-free trajectory planning method. The method comprises the following steps that: (1) the kinematics modelof robots is established; (2) the obverse and inverse kinematics of the robots is analyzed; (3) the Jacobi matrix of the robots is solved; (4) model simplification is performed on the mechanical armsof the robots based on swept ball volumes; (5) a robot collision detection matrix is established, and closest distances between the simplified robot models and obstacles are obtained; (6) the three vector field functions, namely, attraction velocity, repulsion velocity and tangential velocity of a virtual velocity vector field are improved, so that multi-robot collision-free motion planning is performed; and (7) the velocities of the robots are optimized based on a path priority strategy. According to the virtual velocity vector field-based robot three-dimensional collision-free trajectory planning method of the invention, a robot collision-free trajectory planning algorithm is improved, and therefore, the stability of the robots during operation, the smoothness of the trajectories of therobots and the operating efficiency of the robots can be improved.

Virtual velocity vector field-based robot three-dimensional collision-free trajectory planning method

The invention discloses a six-freedom-degree locating method and system used in the process of grabbing objects by an industrial robot. According to the scheme, natural features and local feature points on the surfaces of the objects to be grabbed can be extracted from images, and then recognizing and locating can be carried out through the features. According to the method, next appropriate movement of a robot can be calculated according to joint angles of the robot and a kinematic model and can be executed by the robot. Finally, according to the movement process of the robot and the recognition results of two times, 3D coordinates of vertex points on the surfaces of the objects can be calculated, a corresponding coordinate system of the objects can be established, six-freedom-degree poses of the objects can be obtained, and therefore the corresponding objects can be accurately grabbed.

Six-freedom-degree locating method and system used in process of grabbing objects by industrial robot

The invention provides a dynamic obstacle avoiding system for a capturing process of an industrial robot and a method of the dynamic obstacle avoiding system. The dynamic obstacle avoiding system comprises a rejection vector module, an eddy algorithm module and a dynamic obstacle avoiding algorithm module. The rejection vector module is used for establishing a relative motion relationship between an end effector and an obstacle to obtain a rejection vector; the eddy algorithm module is used for updating the direction of the rejection vector according to an eddy algorithm; and the dynamic obstacle avoiding algorithm module is used for acquiring the position of the end effector according to the rejection vector so as to avoid the obstacle. By the dynamic obstacle avoiding algorithm, the industrial robot successfully avoids the obstacle and finishes capturing under the condition that the moving obstacle exists in an industrial complex environment.

Dynamic obstacle avoiding system for capturing process of industrial robot and method of dynamic obstacle avoiding system

The invention discloses a three-stage convolutional neural network-based object grabbing detection method. The method comprises the following steps of: 1, obtaining a data set; 2, constructing networkstructures of first-stage, second-stage and third-stage convolutional neural networks, and training the convolutional neural networks; 3, obtaining pre-selected grabbing boxes of a target object andjudging values of the pre-selected grabbing boxes by utilizing the trained three-stage serial convolutional neural network; 4, obtaining an optimal grabbing box through the judging values; and 5, determining a position and a posture of the target object. The method is capable of improving the correctness of the grabbing boxes and realizing high-correctness grabbing for unknown objects.

Three-stage convolutional neural network-based object grabbing detection method

The invention discloses a motion planning system and method of an industrial robot for smoothly grabbing a moving object The motion planning system of the industrial robot for smoothly grabbing the moving object comprises a smooth grab model building module, an obstacle avoidance algorithm module and a smooth grab algorithm module The smooth grab model building module is used for describing the relative relationship between the target position and an end actuator and the relationship between an obstacle and the arm plane of the industrial robot so as to generate an attraction vector and a rejection vector The obstacle avoidance algorithm module is used for generating the motion for avoiding the obstacle so as to avoid the moving obstacle The smooth grab algorithm module is used for generating the smooth grab motion according to a smooth grab algorithm in the mode of using the attraction vector and the speed of the moving object as the input under the condition that the obstacle leaves from a collision-free region, so that the smooth grab operation is achieved By the adoption of the motion planning system and method of the industrial robot for smoothly grabbing the moving object, the purpose that the industrial robot smoothly grabs the moving object in the working space in the complex industrial environment can be achieved, and therefore the damage to the object to be grabbed is avoided

Motion planning system and method of industrial robot for smoothly grabbing moving object

In order to realize the object grasping of robotic manipulators in dynamic environments, the moving obstacle should be detected and avoided. When the moving speed of the obstacle is lower than that of the end-effector, a collision avoidance algorithm is proposed by using the repulsive vector to ensure the end-effector far away from the obstacle. When the speed of the obstacle is higher than that of the end-effector, the repulsive vector is further improved with the pivot algorithm. Several experiments are implemented on a 5-DOF robotic manipulator, and the end-effector can successfully grasp the object in the environment with moving obstacle.

Moving obstacle avoidance of a 5-DOF robot manipulator by using repulsive vector

Popular Tasks used by Researchers

How can I help with your research?

Get more enhanced response by uploading the PDFs you want me to reference.

No relevant PDFs in your library

SciSpace is the AI research assistant for academics. Run systematic literature reviews on 280M+ papers, and write papers with cited sources. Trusted by 1M+ students, PhDs & researchers.

SciSpace | AI for Research

We encountered an error while loading the panel. Please try refreshing the page.

Something went wrong

Analyze PDFs

Code & Manuscripts

Funding & Grants

Literature & Patents

Medical & Clinical Data

Systematic Review

Visualize & Present

Web & Data

Build a Google Scholar-like website for your research.

Build a website

Create charts and images for your research

Create a Chart

Write a paper for submission to a journal

Draft a manuscript

Patent Search

Design eye-catching scientific posters in minutes.

Scientific Poster Generation

Systematic Literature Review

Anyone with the link can view this chat. You can switch it back to private anytime.

One task is running at the moment. Your messages will be shown right after.

Drag and drop or click here to browse

Loved by <highlight>1 million+</highlight> researchers

Extract a list of specific topics and their sources from unstructured text

Topics

Compare and analyze relevant papers that matches with your search

Papers

Get insights from PDFs and bookmarked papers from your library

My library

Recent searches

Try searching for:

Catch AI-generated content in scholarly and non-scholarly content

{ai} Detector

Ai Writer

Get PDF Summaries, highlighted text explanations 

Chat with PDF

Effortlessly create in-text citations and bibliographies in APA and 2,500 other formats

Citation generator

Get explanations, summaries, and answers on academic papers

Ease up your research workflow with {scispace}'s cohort of exciting AI tools

Elevate your academic writing skills and convey your ideas the way you want

Paraphraser

Explore our range of reading and writing tools

Your file is being prepared and should be ready in a few minutes. If it's a large file, it might take a bit longer. You can close this window, and we'll email you the file when it's done.

You have reached a maximum limit of <strong>{limit}</strong> columns in the table. Remove at least <strong>1</strong> column to add or create another one.

Zhang Chi

Author Tools

Chat about Author