Vehicle Trajectory Prediction Using Lane And Interaction Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in predicting future trajectories due to environmental complexity and uncertainty, particularly in estimating driver intentions, which affects their navigation and interaction with surrounding vehicles.
Innovation Solution
A system and method that utilize a processor to determine trajectory features and lane features within a target vehicle's vicinity, employing neural networks to predict future locations by encoding position and velocity histories, and using convolutional neural networks to identify probabilities of lane entry, incorporating interaction features from nearby vehicles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional trajectory prediction methods are used, then the system is simpler, but prediction accuracy deteriorates due to environmental complexity and uncertainty
Solution Approach 1:
The system segments the trajectory prediction task into multiple independent modules: a trajectory prediction module that processes historical trajectory data, a lane detection module that identifies road lanes, and an interaction prediction module that analyzes vehicle-vehicle and vehicle-pedestrian interactions. Each module specializes in one aspect, improving overall prediction accuracy while managing complexity through modular design
Solution Approach 2:
The system introduces an interaction feature extraction module as an intermediary between raw sensor data and final trajectory predictions. This intermediary module processes and synthesizes information about interactions between multiple vehicles and pedestrians, enabling more accurate predictions without overwhelming the core prediction algorithm
2Measurement precision
If driver intention is directly estimated, then trajectory prediction becomes more accurate, but the difficulty of detecting and measuring increases
Solution Approach 1:
The system extracts driver intention indirectly by detecting observable behavioral patterns from historical trajectory data, such as lane change tendencies, turning preferences, and response to interactions with other road users. Instead of directly measuring unobservable intention, the system extracts measurable behavioral features that correlate with driver intent
Solution Approach 2:
The system performs preliminary analysis of historical trajectory data and interaction patterns before making predictions. By pre-processing data to extract meaningful features about driver behavior patterns and interaction responses, the system prepares intention indicators in advance, making the actual prediction process more accurate and efficient
3Measurement precision
If more environmental factors are considered, then prediction accuracy improves, but computational complexity increases
Solution Approach 1:
The system applies local quality by focusing computational resources on the most relevant factors for each specific prediction scenario. The interaction prediction module selectively processes information about nearby vehicles and pedestrians based on their relevance to the target vehicle's trajectory, rather than uniformly processing all environmental data
Solution Approach 2:
The system implements partial action by considering only the most significant interaction factors and environmental features for each prediction, rather than exhaustively analyzing all possible factors. This selective approach maintains accuracy while reducing computational burden
Data Source
AI summary
Systems and methods for predicting a location of a target vehicle are disclosed. A processor receives trajectory information for the target vehicle and determines a trajectory feature based on the trajectory information. The processor further determines a lane feature of a lane within a threshold vicinity of the target vehicle, and determines a probability associated with the lane based on the trajectory feature and the lane feature. The lane that the vehicle may enter is identified based on the probability. The processor may also generate interactive features of interactions between the target vehicle and other vehicles. The processor predicts a location of the target vehicle in an upcoming time period based on the lane feature and the trajectory feature. In some embodiments, the interactive features are also used for the prediction.