Manoeuvre Generation for Automated Driving on One-Way Roads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current automated driving systems on one-way roads face challenges in efficiently planning collision-free trajectories while balancing computational resources and optimality, particularly due to non-linear and non-convex collision avoidance constraints.
Innovation Solution
The implementation of model predictive control and quadratic program formulation to determine collision-free reference trajectories for longitudinal and lateral movement, utilizing structured environment constraints to linearly formulate collision avoidance, and integrating with advanced driver assistance systems like adaptive cruise control and lane keeping aid.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If grid/graph based search methods are used for obstacle avoidance, then collision free trajectories can be found, but significant computer memory is required and algorithm efficiency is heavily dependent on heuristics
Solution Approach 1:
The patent transforms the obstacle avoidance problem from a general non-linear optimization problem into a linear quadratic program by changing the parameterization of the cost function and constraints. This allows the use of efficient QP solvers that require minimal memory while guaranteeing collision-free trajectories through linear constraints on vehicle position and velocity.
Solution Approach 2:
The patent replaces complex grid-based search algorithms with a mathematical optimization approach using quadratic programming. This substitution eliminates the need for extensive memory storage of grid cells and heuristic evaluations, while providing a more computationally efficient solution through closed-form or iterative QP methods.
2Reliability
If cost functions or constraints based on distance and relative velocity are used, then collision free trajectories can be achieved, but the cost terms are normally non-linear and/or non-convex providing no guarantee of generating an optimal solution
Solution Approach 1:
The patent reformulates the non-linear cost function into a quadratic form with linear constraints, transforming the optimization problem into a convex quadratic program. This ensures that any local minimum is also a global minimum, providing a guarantee of optimal solution while maintaining collision-free trajectory through linear safety constraints.
3Reliability
If non-linear and non-convex cost terms are used for obstacle avoidance, then collision free trajectories can be achieved, but there is no guarantee of generating an optimal solution and computational resources increase
Solution Approach 1:
The patent substitutes non-linear optimization methods with linear quadratic programming, enabling the use of efficient and well-established QP algorithms. This substitution dramatically improves computational efficiency and productivity while maintaining the reliability of collision-free trajectory generation through linear constraints.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method and arrangement for manoeuvre generation for automated driving on a one-way road of a vehicle (e) hosting the arrangement, the host vehicle (e) further comprising one or more advanced driver assistance systems. A collision free reference trajectory for longitudinal and lateral movement of the host vehicle (e) is determined repeatedly, with a fix time interval, for allowing the host vehicle (e) to retain a desired velocity and maintain the host vehicle (e) in a preferred lane utilizing model predictive control and quadratic program formulation. a control scheme where the structured environment of the one-way road is exploited in order to linearly formulate collision avoidance constraints is utilized. Manoeuvre generation is performed through solving the quadratic program. The one or more advanced driver assistance systems are controlled to perform the generated manoeuvre.