Non-Circular Robot Path Planning With 3D Costmap Refinement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional path planning algorithms for mobile non-circular robots in environments with obstacles are computationally intensive and not suitable for real-time navigation, leading to increased planning time and resource usage.
Innovation Solution
A method for path planning that utilizes a 3D costmap to iteratively select collision-free positions for a mobile non-circular robot, considering its orientation and movement stress, allowing for real-time and efficient collision-free path optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional path planning algorithms (A*, D*) are used for non-circular robots, then the robot can navigate environments with obstacles, but the computational effort and planning time are significantly increased
Solution Approach 1:
The patent segments the path planning problem into two distinct phases: a global planning phase that computes a coarse path using conventional algorithms, and a local refinement phase that optimizes the path for non-circular robot constraints. This segmentation reduces the computational burden on any single algorithm while ensuring both collision-free navigation and efficiency.
Solution Approach 2:
The patent performs preliminary path planning using conventional algorithms to generate an initial feasible path before applying non-circular robot-specific optimizations. This preliminary action provides a starting solution that guarantees obstacle avoidance, which is then refined to reduce planning time for real-time applications.
2Reliability
If conventional path planning algorithms are used for non-circular robots, then collision-free paths can be found, but resource usage increases significantly
Solution Approach 1:
The patent divides the computational workload into global and local planning segments, where the computationally intensive global pathfinding is performed once, and the resource-intensive non-circular robot optimizations are applied only to the specific path segments that require refinement. This reduces overall computational resource consumption.
Solution Approach 2:
The patent applies non-circular robot optimization only to the portions of the path that benefit from it, rather than optimizing the entire path globally. This partial action approach reduces computational resource usage while still achieving collision-free navigation where it matters most.
3Measurement precision
If orientation-dependent distance calculations are performed for non-circular robots, then accurate collision avoidance is achieved, but the complexity of path planning increases
Solution Approach 1:
The patent segments the orientation calculation into a preliminary global phase (using simplified distance metrics) and a local refinement phase (using precise orientation-dependent calculations). This segmentation maintains measurement precision where needed while reducing overall planning complexity.
Solution Approach 2:
The patent performs preliminary path planning using simplified distance calculations that do not require orientation information. Only after a feasible path is established does the system apply the more complex orientation-dependent distance calculations for local refinement, reducing overall computational complexity.
4Reliability
If high spatial resolution path planning is implemented for non-circular robots, then collision-free navigation is ensured, but planning time increases
Solution Approach 1:
The patent applies high spatial resolution optimization only to local path segments where the non-circular robot shape critically affects collision avoidance, rather than applying it uniformly across the entire path. This segmentation ensures collision-free navigation in critical areas while reducing planning time overall.
Solution Approach 2:
The patent implements local quality optimization by applying high-resolution non-circular robot constraints only to specific portions of the path where they are most needed, rather than uniformly across the entire path. This reduces planning time while maintaining collision-free navigation where it matters most.
Data Source
AI summary
A received 3D costmap includes a 2D map of a plane and a cost value per position on the 2D map, wherein the cost value is indicative of a cost of positioning the robot with an orientation in the presence of obstacles in a vicinity of the position. A received start position and end position of the robot include a start and an end orientation of the robot, respectively. Consecutive positions for a path in planning are iteratively selected from the start position to the end position. A cumulative cost value of the consecutive positions, which comprises a sum of the cost values, is minimized. The cumulative cost value comprises a stress value indicative of a change in a movement direction along the planned path.