Autonomous Vehicle Path Prediction for Intersection Collision Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automatic-driving vehicles face the risk of collision due to their inability to sense the future traveling intentions of vehicles in other roads, relying solely on current position data from sensors.
Innovation Solution
A vehicle traveling control method that predicts the traveling paths of candidate vehicles within the sensing range, identifies a target competing vehicle with path intersections, determines traveling strategy combinations based on preset variations and current status, and selects a strategy that meets safe traveling conditions to control the current vehicle's movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensors are used to detect current positions of traveling vehicles, then the current position information can be obtained, but the future traveling intentions cannot be sensed, leading to collision risk
Solution Approach 1:
The system performs preliminary action by predicting future traveling paths and intentions of other vehicles before actual collision risk materializes. The prediction module forecasts multiple possible trajectories based on current sensor data, historical behavior, and road context, enabling the autonomous vehicle to proactively plan avoidance maneuvers rather than reactively responding to immediate threats.
Solution Approach 2:
The patent introduces an intermediary prediction module that bridges the gap between current sensor measurements and future collision risks. This intermediary component processes current position data through probabilistic path prediction algorithms, generating anticipated future states that mediate between present detection capabilities and future safety requirements, allowing the system to 'sense' intentions that are not directly observable.
2Reliability
If multiple traveling strategy combinations are evaluated to ensure safety, then collision avoidance improves, but computational complexity and decision time increase
Solution Approach 1:
The control system is segmented into distinct functional modules: a prediction module that forecasts other vehicles' paths, a strategy generation module that creates multiple candidate maneuvers, an evaluation module that assesses each strategy against safety criteria, and an execution module that implements the selected strategy. This segmentation allows complex safety evaluation to be broken down into manageable computational tasks that can be processed in real-time.
Solution Approach 2:
The system generates and evaluates multiple traveling strategy combinations (excessive action) to ensure comprehensive safety coverage, rather than relying on a single optimal path. By considering more strategies than strictly necessary and evaluating them against preset safe traveling conditions, the system guarantees safety even if some predictions prove inaccurate, accepting the computational overhead as a necessary trade-off for robust safety assurance.
Data Source
AI summary
A vehicle traveling control method, includes: predicting traveling paths of candidate vehicles within the sensing range of a current vehicle; determining a target competing vehicle competing with the current vehicle from the candidate vehicles, where the target competing vehicle is a vehicle having a path intersection with the traveling path of the current vehicle; determining the traveling strategy combinations of the current vehicle and the target competing vehicle according to preset traveling variations and current traveling status of the current vehicle and the target competing vehicle; and determining a target traveling strategy combination that meets preset safe traveling conditions from the traveling strategy combinations, and controlling the traveling of the current vehicle according to the target traveling strategy combination.