Cab-Trailer Path Planning for Trailer Collision Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing path planning algorithms for robots towing trailers fail to account for the trailer's movement during turns, leading to potential collisions with obstacles, and are computationally expensive, especially when dealing with dynamic obstacles.
Innovation Solution
A system and method for forward path planning that estimates the trailer's trajectory based on the robot's movement, using a model-predictive control scheme and the Time Elastic Band planner to insert obstacle constraints and kinematic constraints, while assuming the trailer transitions quickly between steady-state modes, allowing for real-time collision avoidance without direct sensing of the trailer's location.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If high resolution A-star grid search algorithms are used for path planning, then collision avoidance accuracy is improved, but computational speed deteriorates making real-time operation impossible
Solution Approach 1:
The path planning problem is segmented into two parts: global path planning using A-star algorithm on a coarse grid to find the overall route, and local path adjustment using trajectory estimation to avoid dynamic obstacles in real-time. This segmentation allows each part to use appropriate computational resources.
Solution Approach 2:
The system performs preliminary global path planning using A-star algorithm before real-time operation, establishing a baseline path. During real-time operation, only local adjustments are made based on trajectory estimation, avoiding the need for continuous full-grid searches.
2Productivity
If the robot uses a simple path following algorithm, then computational cost is reduced, but the trailer may collide with obstacles during turns
Solution Approach 1:
The system introduces an intermediary trajectory estimation module that predicts trailer position based on robot motion and trailer kinematics. This intermediary layer allows simple path following to be enhanced with accurate collision prediction without direct sensing of trailer location.
Solution Approach 2:
The system replaces direct mechanical sensing of trailer position with a virtual model-based estimation approach. By using kinematic models and robot motion data, the system substitutes physical sensors with computational estimation, reducing system complexity while maintaining accuracy.
3Measurement precision
If direct sensing of trailer location is implemented, then trajectory accuracy is improved, but system complexity and cost increase
Solution Approach 1:
The system uses the robot's own motion data and the trailer's known kinematic properties to estimate trailer position. The trailer effectively 'serves itself' through passive kinematic modeling without requiring active sensing, achieving self-awareness of position through system dynamics alone.
Data Source
AI summary
A system and method for forward path planning of cab trailer systems is provided. Movement of a robot moving forward is tracked and a trajectory of at least one trailer attached to a back of the robot is estimated separate from a path of the robot. The trajectory of the trailer is based on the movement of the robot. A test for collision of one of the trailers is performed by identifying an obstacle and determining a distance of each trailer from the obstacle based on estimated trajectory. The trajectory of the robot is moved away from the obstacle when the distance fails to satisfy a threshold distance from the obstacle.