Orbit Prediction Accuracy via ML Error Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current orbit prediction methods based solely on physics models are not robust enough to achieve the required accuracy for Space Situational Awareness, particularly in collision avoidance scenarios.
Innovation Solution
A computer-based system that utilizes an iterative physics algorithm and a trajectory error prediction machine learning model to improve orbit prediction accuracy by learning correlations between object states and trajectory errors, enabling the display of updated trajectory notifications for users.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If solely physics-based models are used for orbit prediction, then the system complexity is low, but the prediction accuracy is insufficient for required Space Situational Awareness standards
Solution Approach 1:
The patent merges physics-based orbital propagation models with machine learning error prediction models into a hybrid system. The physics model provides baseline trajectory predictions while the ML model learns and predicts errors from historical data, combining both approaches to achieve higher accuracy than either model alone.
Solution Approach 2:
The patent creates a composite prediction system that combines two different modeling approaches (physics-based and data-driven) into a unified framework. The error correction mechanism integrates ML-predicted error estimates with physics-based predictions to produce a composite trajectory prediction with reduced uncertainty.
2Measurement precision
If machine learning models are added to improve prediction accuracy, then the measurement precision improves, but the ease of operation and implementation becomes more difficult
Solution Approach 1:
The system implements feedback by using the ML model to predict errors based on historical trajectory data, then applying these predicted errors to correct future predictions. The model continuously learns from past performance and adjusts its error predictions to improve accuracy over time.
Solution Approach 2:
The ML model is trained in advance on historical trajectory data to learn error patterns before being deployed for real-time prediction. This preliminary training phase allows the model to capture complex error relationships that would be difficult to model through physics alone, preparing it for accurate error prediction during operation.
Data Source
AI summary
Systems and methods of the present disclosure enable machine learning-based refinement of trajectory predictions using a processor to determine a future trajectory associated with an object state using an iterative physics algorithm. The processor utilizes a trajectory error prediction machine learning model to predict a trajectory error for the future trajectory determined for the object state. The processor determines a pseudo-measurement representative of the trajectory error based at least in part on the trajectory error for the future trajectory and determines a pseudo-measurement noise based at least in part on the pseudo-measurement. The processor determines an updated future trajectory for the future trajectory based on the pseudo-measurement and the pseudo-measurement noise of the trajectory error for the future trajectory, and causes to display a trajectory notification associated with future trajectory on a screen of a user computing device associated with a user.