Geographically-Informed Modeling and Analysis of Platform Attitude Jitter in GF-7 Sub-Meter Stereo Mapping Satellite

Haoran Xia, Xinming Tang, Fan Mo, Junfeng Xie, Xiang LiISPRS Int. J. Geo Inf 2024 · DOI

TL;DR

This paper introduces a novel method to model platform attitude jitter on the GF-7 satellite by correlating vibration data with geographic location. This approach helps understand and mitigate jitter, crucial for maintaining high-quality imagery and accurate mapping from this sub-meter stereo mapping satellite.

What Problem Does This Paper Address?

High-resolution Earth observation satellites like the GF-7 are vital for applications ranging from resource monitoring to fundamental surveying. A stable platform and precise attitude control are paramount for achieving high-quality imagery and accurate geospatial data. However, satellite platforms are susceptible to vibrations (jitter) caused by internal and external factors. This jitter, even if small, can degrade image quality, lead to stitching issues, and ultimately reduce mapping accuracy. While previous studies have investigated satellite jitter, there's a need for a model that specifically links this jitter to the satellite's geographic location, considering its orbital path. This paper aims to fill that gap by developing a geographically informed model for GF-7's platform attitude jitter.

Methods

The authors propose a multi-step methodology to construct the geographically informed vibration model:

  1. Data Screening: Raw gyro and star sensor data are filtered to ensure temporal consistency and platform stability, removing erroneous data points caused by noise, environmental fluctuations, or transmission issues. A key criterion is maintaining a consistent sampling interval.
  2. Composite Processing of Gyro and Star Sensor Data:
    • Gyro Data Processing: Gyro data, prone to drift, is fitted with a fifth-order polynomial to improve accuracy.
  3. Star Sensor Data Processing: Star sensor data, represented by quaternions, is used to calculate the optical axis angle, reflecting rapid attitude changes.
  4. Data Integration: To combine the relative information from star sensors and absolute information from gyros, a proportional scaling method (Equation 7) is used. This normalizes the star sensor data to align with the gyro data's range.
  5. Time-Matching: A crucial step addresses the temporal discrepancy between gyro and star sensor data, likely due to thermal fluctuations and transitions in/out of solar shadow. A time-matching method adjusts the star sensor data to align with the gyro data.
  6. Denoising: A moving average filter (Equation 8) is applied to the composite data to reduce noise and enhance the periodic features.
  7. Trigonometric Function Fitting: The denoised composite data is then fitted using a sine function (Equation 9) to capture the periodic characteristics and extract primary frequency components.
  8. Geographic Location-Dependent Flutter Modeling:
    • Geographic Location Calculation: The satellite's geographic location (latitude and longitude) is determined by correlating its orbital data with Beijing time.
  9. Correlation and Modeling: The fitted vibration data (chatter frequency and amplitude) is then correlated with the satellite's geographic location (specifically latitude). Trigonometric functions are used to represent this relationship, leading to a geographically correlated vibration model (Equation 14).

Datasets

The study utilizes original satellite gyro data and star sensor data from the GF-7 satellite. Specific orbital data used for analysis include orbit IDs 016396, 016416, 016431, and 016446, collected between October 12, 2022, and October 15, 2024. The data from orbit ID 016431 (October 14, 2022) was selected for detailed jitter analysis and model construction. No new benchmark datasets were introduced; the study relies on existing satellite operational data.

Key Insights and Ideas

  • Synergy of Gyro and Star Sensors: Combining gyro and star sensor data is essential for accurate jitter detection. Gyros provide continuous, high-frequency relative attitude data, while star sensors offer precise, albeit lower-frequency, absolute attitude information. Their integration, with proper scaling and time-matching, yields a more robust jitter signal.
  • Geographic Correlation is Key: The paper highlights that satellite jitter is not random but exhibits a strong correlation with the satellite's geographic location along its orbit. This is a significant insight, as it suggests that factors related to Earth's gravity field and orbital position play a role in inducing vibrations.
  • Orbital Periodicity: The jitter exhibits a clear periodic behavior that aligns with the satellite's orbital period (approximately 1.5 hours). This periodicity is a fundamental characteristic that can be leveraged for prediction and mitigation.
  • Latitude-Dependent Vibration: The study reveals distinct patterns of vibration amplitude based on latitude and orbital phase (ascending vs. descending). This granular understanding is crucial for targeted operational adjustments.

Key Results

  • Waveform Convergence: A significant convergence in waveform similarities was observed between the gyro data and star sensor optical axis angles, confirming their strong correlation in jitter information.
  • Jitter Amplitude: The recorded vibration amplitude was approximately 2.88".
  • Vibration Period: The jitter period corresponds to the satellite's complete operational cycle, approximately 1.5 hours.
  • Maximum Vibration: The GF-7 satellite experiences maximum vibration amplitude between 5° S and 20° S latitude during its ascending phase.
  • Minimum Vibration: The minimum vibration amplitude is observed between 5° N and 20° N latitude during the descending phase.
  • Model Effectiveness: The developed geographically correlated vibration model effectively captures the relationship between satellite jitter and geographic location, providing a basis for optimizing satellite attitude and mitigating vibrations.

Limitations and Future Work

  • Limitations:
    • The paper acknowledges a temporal discrepancy between gyro and star sensor data, which, despite time-matching efforts, might still affect the accuracy of composite data. This is attributed to factors like thermal fluctuations and the onboard thermal control system's limitations.
  • While the model is developed, its direct application for real-time vibration control or optimization is not detailed.
  • Future Work:
    • The authors suggest further research to delve deeper into the underlying causes of the observed vibrations, potentially exploring factors like Earth's gravity field distribution and orbital elevation changes.
  • Future work will also involve conducting annual cycle frequency analyses, comprehensively analyzing the jitter phenomenon, and refining the vibration model.
  • The study aims to provide a reference for optimizing satellite operation, especially for vibration control in specific orbital segments, to further improve remote-sensing data quality.

Summary updated

BibTeX
@article{xia2024geographically,
  author = {Haoran Xia and Xinming Tang and Fan Mo and Junfeng Xie and Xiang Li},
  title = {{Geographically-Informed Modeling and Analysis of Platform Attitude Jitter in GF-7 Sub-Meter Stereo Mapping Satellite}},
  journal = {ISPRS Int. J. Geo Inf.},
  year = {2024},
  doi = {10.3390/ijgi13110413},
}
Figure