Unmanned Vehicle Path Smoothing With Consistent Obstacle Detours
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Unmanned vehicles face inconsistent detour behaviors when smoothing paths due to lack of lateral decisions for obstacles, leading to ineffective obstacle avoidance.
Innovation Solution
A method for optimizing vehicle paths by determining whether to detour obstacles left or right based on the relative position of adjacent waypoints and the obstacle's center point, using a Frenet coordinate system to analyze obstacle boundaries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Shape
If path smoothing processing is applied to the initial path, then the path smoothness is improved, but the detour behavior consistency deteriorates because there is no lateral decision for obstacles around the path
Solution Approach 1:
The patent applies preliminary action by making lateral decisions for obstacles before path smoothing is performed. The system evaluates obstacles around the initial path and determines detour directions (left, right, or no detour) in advance, then applies smoothing while preserving these predetermined lateral decisions. This ensures that the smoothed path maintains consistent detour behavior with the original path.
2Device complexity
If the optimal path is obtained based on the evaluation system without lateral decisions for obstacles, then the path planning complexity is reduced, but the obstacle avoidance effectiveness deteriorates
Solution Approach 1:
The patent applies segmentation by dividing the path planning process into two independent parts: (1) obtaining the optimal path based on the evaluation system without considering lateral decisions, and (2) making lateral decisions for obstacles around this path. This segmentation allows the system to maintain low planning complexity while improving obstacle avoidance effectiveness through the additional lateral decision-making step.
Solution Approach 2:
The patent applies dimensionality change by adding a lateral dimension to the path planning problem. Instead of only optimizing the longitudinal path based on the evaluation system, the system introduces lateral decisions (left, right, no detour) as an additional dimension of optimization, enabling the vehicle to effectively avoid obstacles while maintaining the simplicity of the original path planning approach.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
An unmanned vehicle path optimization method and apparatus, a computer readable storage medium, and an electronic device, relating to the technical field of computers and communications. The method comprises: obtaining a generated initial path of an unmanned vehicle (S510); obtaining an obstacle in the periphery of the initial path (S520); if the unmanned vehicle intersects with the obstacle on the initial path, deciding not to detour the obstacle (S530); if a second waypoint is located on the left side of a line connecting a first waypoint and a central point of the obstacle, deciding to detour the obstacle leftwards; if the second waypoint is located on the right side of the line connecting the first waypoint and the central point of the obstacle, deciding to detour the obstacle rightwards (S540); and optimizing the initial path of the unmanned vehicle on the basis of the decision for the obstacle (S550). The present method can solve the problem that detouring behaviors of a smoothed path and an unsmoothed path on an obstacle are not uniform.