Vehicle Trajectory Prediction via Server-Client Feedback Loop
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional track prediction methods for moving objects, such as vehicles and pedestrians, lack accuracy due to the failure to consider dynamic operations like driver actions and pedestrian movements, which significantly impact future trajectories.
Innovation Solution
A track prediction method and device that involves a server-side prediction of a moving object's trajectory based on current and historical data, including driver operations like acceleration and braking, which is then iteratively refined by the moving object and sent back to the server, utilizing high computing capabilities and precise movement states.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional track prediction methods using historical trajectories and models are used, then the prediction process is simple, but the accuracy of track prediction is low
Solution Approach 1:
The patent implements a feedback mechanism where the moving object receives the predicted track from the server, compares it with actual operations (acceleration, braking, steering), and sends correction information back to the server. This iterative feedback loop continuously refines the prediction model, improving accuracy over time while managing complexity through structured data exchange protocols.
Solution Approach 2:
The system transitions from static historical trajectory analysis to dynamic real-time prediction by incorporating live operation data. The prediction model dynamically adjusts based on current acceleration, braking, and steering inputs, enabling accurate prediction of future trajectories under changing conditions rather than relying solely on past patterns.
2Measurement precision
If operations associated with moving objects are considered in track prediction, then track prediction accuracy is improved, but the complexity of data processing increases
Solution Approach 1:
The patent segments the prediction system into distinct functional modules: a server-side prediction module that processes historical and real-time data, and a client-side execution module that handles local operations. This segmentation allows complex data processing to be distributed, with the server managing heavy computational tasks and the moving object handling local operation data, thereby reducing overall system complexity.
Solution Approach 2:
The system changes the parameters used in prediction from purely historical trajectory data to include real-time operation parameters such as acceleration, braking force, and steering angle. By incorporating these dynamic operation parameters alongside traditional position and velocity data, the model achieves higher accuracy while the structured approach to parameter integration manages processing complexity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention relates to a track prediction method for a moving object, the method comprising: receiving a first predicted track of the moving object from a server side; changing the first predicted track at least based on an operation associated with the moving object, so as to acquire a second predicted track; and sending the second predicted track to the server side. The present invention further relates to a track prediction method for a server side for communicating with a moving object, a track prediction device for a moving object, a track prediction device for a server side for communicating with a moving object, an automobile, a server, and a computer storage medium. With a track prediction solution provided by the present invention, an operation associated with the moving object (for example, an operation performed by a driver or a movement performed by a pedestrian) is considered in track prediction, so that a more accurate final track prediction result may be obtained.