Obstacle Trajectory Prediction Using Historical Interaction Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current obstacle trajectory prediction methods for unmanned vehicles rely solely on current status information, which limits the accuracy of predicting future motion trajectories of moving obstacles, as they do not adequately consider the influence of historical interactions between the vehicle and obstacles.
Innovation Solution
The method determines a current interaction feature using historical and current status information, then uses this to calculate a global interaction feature incorporating the vehicle's planned future motion trajectory, enhancing the reliability of predicting future obstacle trajectories by considering priori knowledge of future interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If only current status information is used for obstacle trajectory prediction, then the prediction process is simple and fast, but the prediction accuracy is limited
Solution Approach 1:
The patent applies preliminary action by pre-training a trajectory prediction model offline using historical interaction data between vehicles and obstacles. This pre-trained model captures complex interaction patterns in advance, allowing the online prediction system to achieve high accuracy without real-time computational complexity. The model is trained beforehand with various interaction scenarios, enabling it to make accurate predictions when deployed in real-time vehicle operation.
2Reliability
If historical interaction data is integrated into the prediction model, then the reliability of predicting future obstacle trajectories improves, but the computational complexity increases
Solution Approach 1:
The patent resolves this contradiction by performing the complex computation of learning interaction patterns in advance during offline training. The pre-trained model encodes historical interaction data between vehicles and obstacles into learned parameters and features. During real-time prediction, the model simply applies these pre-learned patterns to current status information, achieving high reliability without real-time computational burden.
Solution Approach 2:
The patent uses copying by creating a virtual representation of historical interaction patterns through the pre-trained model. Instead of processing actual historical data in real-time, the model copies the essential interaction patterns into its structure during training, then uses these copied patterns for rapid prediction during operation.
3Measurement precision
If the vehicle's planned future motion trajectory is incorporated into interaction feature determination, then the accuracy of representing future interactions improves, but the processing requirements increase
Solution Approach 1:
The patent applies preliminary action by determining the vehicle's future motion trajectory in advance through path planning algorithms. This pre-determined trajectory information is then fed into the interaction feature determination process, allowing the model to accurately represent future vehicle-obstacle interactions without real-time iterative computation. The future trajectory serves as prior knowledge that guides the prediction process.
Data Source
AI summary
An obstacle trajectory prediction method and apparatus are provided. In some embodiments, a global interaction feature under joint action of a vehicle and obstacles is determined according to historical status information and current status information of the vehicle, historical status information and current status information of the obstacles, and a future motion trajectory planned by the vehicle; an individual interaction feature of a to-be-predicted obstacle is determined according to the global interaction feature and current status information of the to-be-predicted obstacle; and a future motion trajectory of the to-be-predicted obstacle is predicted through the individual interaction feature and information about an environment around the vehicle.