Autonomous Satellite Navigation via Landmark Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional spacecraft navigation techniques require assumptions about central body orientation and rely on external corrections, which are problematic for high-accuracy long-duration autonomous flight, especially in deep space missions where errors in orientation and time measurement are significant.
Innovation Solution
The implementation of systems and methods using computer vision and machine learning algorithms to autonomously recognize landmarks, allowing spacecraft to estimate Earth's orientation and navigate without GPS or regular ground contacts, using convolutional neural networks, image segmentation, and algorithms like SIFT and ORB for key point detection and matching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If conventional navigation techniques are used with assumptions about central body orientation, then the navigation system is simpler to implement, but the accuracy deteriorates for long-duration autonomous flight
Solution Approach 1:
The spacecraft performs self-calibration of its orientation reference frame by autonomously recognizing landmarks and computing the transformation between the inertial frame and the body-fixed frame. This eliminates the need for external corrections and assumptions about central body orientation, allowing the system to maintain high accuracy without continuous ground support.
Solution Approach 2:
The system pre-loads a database of landmark images and recognition algorithms before the mission. This preliminary preparation enables the spacecraft to perform autonomous landmark recognition and orientation determination without requiring real-time external assistance, resolving the contradiction between system simplicity and long-duration accuracy.
2Measurement precision
If external corrections from ground stations are obtained, then the navigation accuracy is improved, but the autonomy of the spacecraft deteriorates
Solution Approach 1:
The spacecraft autonomously performs landmark recognition, computes the inertial positions of landmarks, and determines the transformation between reference frames without requiring external corrections from ground stations. This self-service capability achieves both high navigation accuracy and complete autonomy for long-duration missions.
Solution Approach 2:
The system uses onboard sensors to continuously monitor landmark positions and compares them against the pre-loaded database, creating a closed-loop feedback system that maintains navigation accuracy autonomously without external intervention.
3Measurement precision
If known landmarks are used for navigation, then the sensitivity for precise orbit determination is improved, but the requirement for landmark identification in various lighting conditions increases complexity
Solution Approach 1:
The system pre-loads a comprehensive database of landmark images captured under various lighting conditions, angles, and resolutions before the mission. This preliminary preparation enables the spacecraft to recognize landmarks autonomously without requiring complex real-time identification algorithms, resolving the contradiction between precision and complexity.
Solution Approach 2:
The system uses multiple copies of landmark images taken from different perspectives, lighting conditions, and distances in the database. This allows the recognition algorithm to match landmarks under varying conditions by comparing against multiple pre-captured copies, maintaining high sensitivity without increasing operational complexity.
4Device complexity
If unknown landmarks are used for navigation, then the requirement for landmark databases and identification systems is reduced, but the sensitivity for precise orbit determination deteriorates
Solution Approach 1:
The system uses a universal database of landmark images that can match landmarks regardless of lighting conditions, viewing angles, or distance. This multi-functional approach allows the same database to serve both known and unknown landmarks, maintaining high orbit determination sensitivity while avoiding the need for separate identification systems.
Data Source
AI summary
Systems and methods are provided for high fidelity long-duration autonomous spacecraft navigation relative to a planet's surface and measuring the dynamics of the planet. For a planet like Earth, embodiments of the present disclosure can be used to estimate the unpredictable components of Earth's orientation with respect to the inertial frame. Embodiments of the present disclosure further enable autonomous landmark navigation by providing systems and methods for satellites to autonomously recognize landmarks, using, for example, multiple computer vision approaches to recognize multiple types of landmarks.