INS/CNS Deeply Integrated Navigation Method of Near Space Vehicles

Rongjun Mu, Hongchi Sun, Yuntian Li, Naigang CuiSensors 2020 · DOI · OpenReview

TL;DR

This paper introduces a novel INS/CNS (Inertial Navigation System/Celestial Navigation System) deeply integrated navigation method for near-space vehicles. It addresses the challenges of long-term accuracy, dynamic response, and aero-optical effects by proposing a new integration model and a robust H-infinity filter, achieving significant improvements in computational efficiency and attitude accuracy.

What Problem Does This Paper Address?

Near-space vehicles, operating in the 20-100 km altitude range, demand highly accurate navigation systems. While celestial navigation offers long-term accuracy by referencing fixed stars, traditional methods suffer from slow star identification, limiting dynamic response. Furthermore, the near-space environment introduces aero-optical effects that corrupt sensor measurements with colored noise, degrading the performance of standard navigation filters. Existing integrated navigation systems (loosely and tightly coupled) also face limitations in computational efficiency and robustness against these environmental factors. This paper aims to overcome these challenges by developing a more efficient and robust navigation solution for near-space vehicles.

Methods

The paper proposes two main innovations:

  1. INS/CNS Deeply Integrated Navigation Model: This model goes beyond traditional loose and tight integration by incorporating INS data directly into the star sensor's image processing. Instead of relying on explicit star identification, it optimizes a "gray image error function" derived from the difference between the predicted star image (from INS) and the observed star image. This approach significantly reduces computational cost and allows for navigation even with fewer than three visible stars.
  2. Second-Order State Augmented H-Infinity Filter: To combat the colored noise introduced by aero-optical effects, a novel H-infinity filter is developed. This filter augments the state vector to model the colored noise as a second-order Markov process. This state augmentation, combined with the inherent robustness of H-infinity filtering against model uncertainties, effectively whitens the measurement noise and improves estimation accuracy.

Datasets

The paper uses the Smithsonian Astrophysical Observatory (SAO) catalog for simulation validation. This catalog is a standard reference for star positions. The simulations were conducted under three working conditions: low, medium, and high uncertainty, to assess the algorithm's robustness.

Key Insights and Ideas

  • Deep Integration for Efficiency: By directly using image data and INS predictions, the need for computationally expensive star identification is eliminated, leading to substantial speedups.
  • Robustness to Aero-Optical Effects: The colored noise introduced by aero-optical effects is a major hurdle. The proposed second-order state augmented H-infinity filter effectively handles this by modeling and whitening the noise, a critical step for reliable navigation in near-space.
  • Unmatched Stars Don't Derail Navigation: A surprising finding is that even with incorrectly matched stars, the deeply integrated model's optimization process remains stable and accurate, as the constant error introduced by unmatched stars doesn't affect the final optimization result.
  • Double Star Sensors are Crucial: For near-space vehicles, the accuracy of the yaw angle is critical for positioning. The study highlights that a single star sensor configuration leads to significant yaw error, recommending a double star sensor setup for improved overall accuracy.

Key Results

  • Computational Cost Reduction: The proposed INS/CNS deeply integrated navigation method reduces computational cost by 50% compared to tightly integrated methods and 60% compared to loosely integrated methods.
  • Attitude Accuracy: The attitude accuracy is maintained within 10" (3 σ) for the non-optical axis and 50" (3 σ) for the optical axis in the deeply integrated mode.
  • Filter Performance: The second-order state augmented H-infinity filter achieves an attitude root mean square error of less than 5" even with a 50% parameter error in the colored noise model.
  • Robustness: The algorithm demonstrated convergence and maintained accuracy even under high uncertainty conditions where the INS only correctly predicted 30% of navigation stars.

Limitations and Future Work

  • Initial Filter Error: The second-order state augmented H-infinity filter exhibits a larger initial error (around 3 seconds) due to the unknown initial values of the colored noise states. This error gradually reduces as the filter estimates the colored noise. The authors suggest that results should be used at least 10 seconds after filtering begins.
  • Double Star Sensor Recommendation: While the paper demonstrates the benefits of a double star sensor configuration for improved yaw accuracy, the primary focus of the paper is on the integrated navigation method and filter. Future work could further explore optimal configurations and their impact.
  • Parameter Sensitivity: While the filter is robust to parameter errors in the colored noise model (up to 50%), further investigation into the optimal range and sensitivity of these parameters in real-world scenarios could be beneficial.

Summary updated

Abstract

Celestial navigation is required to improve the long-term accuracy preservation capability of near space vehicles. However, it takes a long time for traditional celestial navigation methods to identify the star map, which limits the improvement of the dynamic response ability. Meanwhile, the aero-optical effects caused by the near space environment can lead to the colorization of measurement noise, which affects the accuracy of the integrated navigation filter. In this paper, an INS/CNS deeply integrated navigation method, which includes a deeply integrated model and a second-order state augmented H-infinity filter, is proposed to solve these problems. The INS/CNS deeply integrated navigation model optimizes the attitude based on the gray image error function, which can estimate the attitude without star identification. The second-order state augmented H-infinity filter uses the state augmentation algorithm to whiten the measurement noise caused by the aero-optical effect, which can effectively improve the estimation accuracy of the H-infinity filter in the near space environment. Simulation results show that the proposed INS/CNS deeply integrated navigation method can reduce the computational cost by 50%, while the attitude accuracy is kept within 10” (3 σ ). The attitude root mean square of the second-order state augmented H-infinity filter does not exceed 5”, even when the parameter error increases to 50%, in the near space environment. Therefore, the INS/CNS deeply integrated navigation method can effectively improve the rapid response ability of the navigation system and the filtering accuracy in the near space environment, providing a reference for the future design of near space vehicle navigation systems.
BibTeX
@article{mu2020ins,
  author = {Rongjun Mu and Hongchi Sun and Yuntian Li and Naigang Cui},
  title = {{INS/CNS Deeply Integrated Navigation Method of Near Space Vehicles}},
  journal = {Sensors},
  year = {2020},
  doi = {10.3390/s20205885},
}

Loading related content...

Figure