Graph-Based Lane Change Guide for Autonomous Vehicles
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional motion planning and control systems for autonomous vehicles do not accurately account for differences in vehicle types, leading to inefficient and potentially unsafe lane changes, as they rely solely on curvature and speed without considering specific vehicle features.
Innovation Solution
A graph-based lane change guide is implemented using a topological graph to determine the earliest and last nodes in the source lane where a lane change is possible, allowing for a range of lane change segments, enabling more precise and efficient lane changes by performing a conditional breadth-first search to identify feasible nodes between the source and target lanes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional motion planning and control systems use only curvature and speed for lane change decisions, then the system is simple to implement, but the lane change accuracy and safety deteriorate because vehicle type differences are not considered
Solution Approach 1:
The patent segments the lane change decision-making process into multiple components: vehicle feature extraction module, topological graph construction module, and lane change guide generation module. This segmentation allows the system to handle complex vehicle type differences through structured processing while maintaining manageable system architecture.
Solution Approach 2:
The patent introduces a topological graph as an intermediary data structure that bridges vehicle features and lane change decisions. The graph encodes spatial relationships and lane connectivity information, serving as a mediator that translates complex environmental and vehicle-specific parameters into actionable lane change guidance without requiring direct complex processing of all input variables.
2Reliability
If conventional systems apply the same motion planning to all vehicle types, then the system is easy to operate, but the reliability of lane changes deteriorates due to lack of vehicle-specific adaptation
Solution Approach 1:
The patent applies local quality by extracting and utilizing specific vehicle features (such as length, width, turning radius) that are relevant to lane change operations. Different vehicle types receive customized lane change guidance based on their specific characteristics, with the system adapting its planning parameters locally for each vehicle type rather than applying uniform rules globally.
Solution Approach 2:
The patent changes key planning parameters based on vehicle type characteristics. The motion planning parameters (such as lateral acceleration limits, lane change duration, safety margins) are dynamically adjusted according to the extracted vehicle features, allowing the system to maintain high reliability across different vehicle types while automating the parameter adjustment process.
3Measurement precision
If the system determines lane change ranges using comprehensive graph search methods, then the lane change guidance becomes more accurate, but the computational time increases
Solution Approach 1:
The patent performs preliminary construction of the topological graph representing the road network and lane connectivity before actual lane change decisions are needed. This pre-processing creates a ready-to-use spatial model that can be quickly queried during real-time operation, avoiding the need to perform comprehensive graph searches from scratch for each lane change decision.
Solution Approach 2:
The patent extracts only the essential topological features and lane connectivity information needed for lane change planning, rather than processing the complete road map data. By taking out and focusing on the relevant subset of spatial data (lane segments, connectivity relationships, and geometric properties), the system achieves accurate lane change range determination with reduced computational overhead.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
In one embodiment, autonomous driving control is provided for an autonomous vehicle changing from a source lane to a target lane. Using a topological map, a reference node is selected in the source lane. With respect to the reference node, an earliest node is determined in the source lane at which it is first possible for the vehicle to change lanes and a last node is determined in the source lane after which it is no longer possible to change lanes. A range of the source lane is determined for which the vehicle can change lanes.