Latent force models in autonomous GNSS satellite orbit prediction

Rautalin, Sakari; Ali-Löytty, Simo; Piché, Robert

doi:10.1109/icl-gnss.2017.8376242

Cited by 5 publications

(2 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By assuming Gaussian process (GP) priors on unknown forces, this method allows to determine future orbit positions and corresponding uncertainty. This work was by extended Rautalin et al [59] who obtained positive results for a set of satellite constellations in Medium Earth Orbit (MEO) and Geo-Synchronous Orbit (GEO).…”

Section: Orbit Predictionmentioning

confidence: 95%

Machine Learning in Orbit Estimation: a Survey

Caldas¹,

Soares²

2022

Preprint

View full text Add to dashboard Cite

Since the late '50s, when the first artificial satellite was launched, the number of resident space objects (RSOs) has steadily increased. It is estimated that around 1 Million objects larger than 1 cm are currently orbiting the Earth, with only 30,000, larger than 10 cm, presently being tracked. To avert a chain reaction of collisions, termed Kessler Syndrome [1], it is indispensable to accurately track and predict space debris and satellites' orbit alike. Current physics-based methods have errors in the order of kilometres for 7 days predictions, which is insufficient when considering space debris that have mostly less than 1 meter. Typically, this failure is due to uncertainty around the state of the space object at the beginning of the trajectory, forecasting errors in environmental conditions such as atmospheric drag, as well as specific unknown characteristics such as mass or geometry of the RSO. Leveraging datadriven techniques, namely machine learning, the orbit prediction accuracy can be enhanced: by deriving unmeasured objects' characteristics, improving non-conservative forces' effects, and by the superior abstraction capacity that Deep Learning models have of modelling highly complex non-linear systems. In this survey, we provide an overview of the current work being done in this field.

show abstract

Section: Orbit Predictionmentioning

confidence: 95%

Machine Learning in Orbit Estimation: a Survey

Caldas¹,

Soares²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…However, simulation results showed that adding the small perturbation forces only slightly improved the accuracy of the prediction. Another study [17] proposed a state-space model for describing the potential forces to address the shortcomings of existing force models, but simulation results showed that this was not always effective for all satellites. In general, because of the complex perturbation factors associated with the space environment, it is extremely difficult to establish accurate physical models to describe all of the forces acting on a satellite.…”

Section: Introductionmentioning

confidence: 99%

Assisted cold start method for GPS receiver with artificial neural network-based satellite orbit prediction

Yang¹,

Zhang²,

Chen³

et al. 2020

Meas. Sci. Technol.

View full text Add to dashboard Cite

Traditional GPS receivers usually take a long time to obtain their first position solution in the cold start mode. This paper presents a method for assisting the cold start of GPS receivers by autonomous satellite orbit prediction. In this method, the time-consuming broadcast ephemeris collection process in a traditional cold start is replaced by orbit prediction to obtain the satellite position information required for positioning. Thus, the time to first fix of the receiver can be reduced significantly. In addition, since the traditional orbit prediction algorithms based solely on physics-based models have limited accuracy, a hybrid model composed of dynamic models and artificial neural network models for orbit prediction is proposed to improve the accuracy of the orbit prediction. Taking GPS PRN1 as an example, the simulation results show that, in 75% of cases, the satellite position error after 7 days’ prediction with the dynamic models is less than 162.84 m, but this value can be reduced to 46.06 m with the proposed hybrid model; the improvement rate achieved is approximately 71.71%. Some positioning experiments with simulated satellite signals were conducted to assess the first-fix accuracy of a GPS receiver with the proposed satellite orbit prediction function. The results of this study offer insights for the design of advanced assisted GPS receivers.

show abstract

Precise and Efficient Orbit Prediction in LEO with Machine Learning using Exogenous Variables

Caldas,

Soares

2024

2024 IEEE Congress on Evolutionary Computation (CEC)

View full text Add to dashboard Cite

Latent force models in autonomous GNSS satellite orbit prediction

Cited by 5 publications

References 8 publications

Machine Learning in Orbit Estimation: a Survey

Machine Learning in Orbit Estimation: a Survey

Assisted cold start method for GPS receiver with artificial neural network-based satellite orbit prediction

Precise and Efficient Orbit Prediction in LEO with Machine Learning using Exogenous Variables

Contact Info

Product

Resources

About