Vehicle Control Device Route Prediction via Modular Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle behavior prediction methods are unable to accurately predict the route of another vehicle, leading to potential inaccuracies in vehicle control and smooth travel.
Innovation Solution
A vehicle control device and method that recognizes the surrounding environment, including road structure and vehicle actions, to derive a predicted probability of the other vehicle's route, prioritizing explicit and implicit actions, and controlling vehicle behavior based on this prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional vehicle behavior prediction methods are used, then the system is simple to implement, but the prediction accuracy of another vehicle's route is insufficient
Solution Approach 1:
The prediction model is segmented into multiple independent components: road structure analysis module, explicit action recognition module (direction indicators), implicit action recognition module (vehicle dynamics), and probability integration module. Each component processes specific aspects of vehicle behavior separately, then combines results to achieve high prediction accuracy while maintaining manageable system complexity through modular design.
Solution Approach 2:
The system transitions from traditional single-dimension prediction (only based on current position) to multi-dimensional prediction by incorporating road structure characteristics, explicit actions (direction indicators), implicit actions (acceleration, deceleration, lane changes), and environmental context. This dimensional expansion significantly improves prediction accuracy by considering multiple factors simultaneously.
2Adaptability or versatility
If multiple routes are assumed for prediction, then the prediction comprehensiveness improves, but the computational complexity increases
Solution Approach 1:
The system performs preliminary classification of possible routes based on road structure analysis before detailed probability calculation. By pre-identifying feasible routes (e.g., straight, left turn, right turn based on intersection type and lane configuration), the system reduces the search space and computational burden while maintaining comprehensive coverage of all plausible trajectories.
Solution Approach 2:
The system calculates predicted probabilities for multiple routes beyond what is strictly necessary, including low-probability scenarios. This excessive action ensures comprehensive coverage of all possible vehicle behaviors, allowing the system to adapt to unexpected maneuvers while maintaining computational efficiency through probability thresholding for final decision-making.
Data Source
AI summary
A vehicle control device includes a recognizer configured to recognize a surrounding environment including a structure of a road near a vehicle and another vehicle, a deriver configured to derive a predicted probability that the other vehicle will travel in the future along each of routes which are assumed when a plurality of routes along which the other vehicle is able to travel are assumed on a road on which the other vehicle recognized by the recognizer travels, and a travel controller configured to control behavior of the vehicle based on the predicted probability derived by the deriver.