Robotic Arm Path Planning with Potential Field Escape From Local Minima
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robotic arm obstacle avoiding technologies face challenges in dynamic environments due to multiple joints, flexibility, and diverse postures, leading to issues like local minima and unreachable goals, with current methods failing to provide a robust real-time obstacle avoidance solution.
Innovation Solution
A method employing virtual joint interpolation, spherical bounding boxes, RRT-Connect algorithm, and an improved artificial potential field with guiding forces and distance influence factors to enhance path planning and avoid local minima and unreachable goals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If real-time planning algorithms are used to accommodate dynamic obstacle changes, then adaptability to dynamic environments is improved, but the robotic arm easily triggers local minima and stops moving
Solution Approach 1:
The patent applies dynamics by making the potential field parameters adjustable and adaptive. The repulsive force coefficient and attractive force coefficient are dynamically adjusted based on the robotic arm's position relative to obstacles and goal, allowing the planning algorithm to adapt to changing environments while maintaining motion continuity through coefficient modulation rather than complete re-planning
Solution Approach 2:
The patent changes parameters by introducing dynamic coefficients (repulsive force coefficient and attractive force coefficient) that modify the artificial potential field characteristics in real-time. These parameter changes allow the system to adapt to dynamic obstacles without getting trapped in local minima, as the field landscape continuously evolves to guide the robotic arm toward the goal
2Measurement precision
If the robotic arm is modeled with multiple joints and flexibility for accurate representation, then modeling precision is improved, but computational complexity increases making real-time planning difficult
Solution Approach 1:
The patent uses simplified spherical models to represent the robotic arm links and end-effectors instead of complex multi-joint geometries. These simplified models are computationally inexpensive to process and sufficient for collision detection purposes, sacrificing detailed geometric accuracy for real-time planning capability while maintaining adequate safety through the artificial potential field approach
3Productivity
If traditional artificial potential field method is used for local path planning, then real-time obstacle avoidance is achieved, but the robotic arm falls into local minima and cannot reach the goal
Solution Approach 1:
The patent implements feedback by continuously monitoring the robotic arm's position, obstacle positions, and goal position to dynamically adjust the repulsive and attractive force coefficients. This feedback mechanism ensures that when the robotic arm approaches local minima, the coefficients are modified to escape the minimum and continue toward the goal, maintaining both real-time responsiveness and goal reachability
Solution Approach 2:
The patent applies periodic action through the continuous recalculation and updating of the artificial potential field at each control cycle. The field is periodically regenerated with updated obstacle positions and adjusted coefficients, creating a time-varying potential landscape that prevents the robotic arm from settling into static local minima and ensures progressive movement toward the goal
Data Source
AI summary
A robotic arm obstacle avoiding path planning method is provided. The method includes the following steps: step 1: simplifying the robotic arm model and obstacles, determining robotic arm joint points, and adopting virtual joint interpolation to interpolate connecting rods between adjacent joints; employing spherical bounding boxes at each joint point to envelop and replace the robotic arm model, enabling complete substitution for distance calculation when the robotic arm assumes any posture; step 2: adopting an eye-to-hand configuration to position the depth camera, acquiring in real time the point cloud information of the robotic arm and obstacles in the workspace, and using a robot real-time filtering package to filter out the point cloud information of the robotic arm itself.