Robot Path Planning via Node-Edge Validity Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot path planning methods, particularly for manipulator robots with high degrees of freedom, face inefficiencies in recalculating link functions between obstacle and sampling spaces, leading to prolonged computation times and unbalanced tree structures, which can result in longer paths and reduced performance when obstacles change dynamically.
Innovation Solution
A method that first checks for overlap between obstacles and nodes, removes overlapping nodes, generates a driving path using remaining nodes and edges, and then checks for edge overlaps, allowing for more efficient path planning by prioritizing node validity checks and selectively updating the robot state tree.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If random sampling-based re-planning is performed in a variable environment, then the robot can adapt to newly recognized obstacles, but the computation time increases significantly due to repeated validity checks
Solution Approach 1:
The validity check process is segmented into two distinct phases: node validity check (checking whether robot states at nodes collide with obstacles) and edge validity check (checking whether the robot collides with obstacles during motion along edges). This segmentation allows the system to perform frequent, lightweight node checks for rapid adaptability, while performing computationally intensive edge checks less frequently, thus resolving the contradiction between adaptability and computation time.
2Reliability
If validity check is performed for all sampling data whenever obstacle is updated, then complete obstacle avoidance is achieved, but the computational efficiency deteriorates
Solution Approach 1:
The patent performs preliminary node validity checks immediately when obstacles are updated, before performing the more computationally intensive edge validity checks. By first identifying and removing invalid nodes, the system reduces the number of edges that need subsequent validation, thereby maintaining reliable obstacle avoidance while improving computational efficiency.
Solution Approach 2:
The system performs node validity checks for all nodes in the configuration space (excessive action) to ensure complete obstacle avoidance, while performing edge validity checks selectively on only those edges connected to invalid nodes or target nodes (partial action). This selective approach maintains reliability while significantly improving productivity.
3Reliability
If the robot state tree is continuously updated with new sampling data, then path planning completeness is improved, but the tree structure becomes unbalanced leading to longer paths
Solution Approach 1:
The patent implements periodic rebalancing of the robot state tree at specific intervals or trigger conditions (such as when a certain number of nodes are added or when computation time exceeds a threshold). This periodic rebalancing prevents the tree from becoming excessively unbalanced, thereby maintaining shorter paths while still allowing continuous updates for path planning completeness.
Data Source
AI summary
A robot control system includes a detecting device configured to detect an obstacle, and a control device configured to control a robot by planning a driving path of the robot. The control device generates a road map space including nodes containing information about a state of the robot and edges connecting the nodes. Also, the control device checks whether the obstacle detected by the detecting device overlaps with the nodes, removes an overlapped node, generates a driving path by sampling remaining nodes and edges, and checks whether the edges of the driving path overlap with the obstacle. When there is no overlapped edge, the control device determines the driving path on the road map space to plan a path along which the robot drives without colliding with the obstacle.