A Novel Ranging and IMU-Based Method for Relative Positioning of Two-MAV Formation in GNSS-Denied Environments

Jian Cheng, Peng Ren, Tingxiang Deng2023 · DOI

TL;DR

This paper introduces a new method for two Micro Aerial Vehicles (MAVs) to precisely determine their relative positions in environments where GPS signals are unavailable. By fusing data from onboard Inertial Measurement Units (IMUs) and inter-vehicle distance measurements, the system achieves robust and accurate positioning without relying on external anchors, significantly outperforming existing methods.

What Problem Does This Paper Address?

Coordinated missions for Micro Aerial Vehicles (MAVs) are becoming increasingly important for tasks like surveillance and environmental monitoring. However, these missions often take place in complex environments (e.g., urban canyons, forests) where Global Navigation Satellite Systems (GNSS) signals are unreliable or completely unavailable. Relying solely on an MAV's Inertial Measurement Unit (IMU) for positioning is problematic due to inherent error accumulation over time. Existing methods for relative positioning in GNSS-denied environments often have limitations, such as requiring reference anchors, suffering from poor robustness in challenging conditions, or being susceptible to interference. This paper aims to address these challenges by developing a robust and accurate relative positioning method for MAV formations that doesn't depend on GNSS or external anchors.

Methods

The core of the proposed method is a cooperative relative positioning system that fuses data from two sources:

  • Inter-vehicle Ranging: MAVs measure the distance between themselves. This provides a direct measurement of their relative separation.
  • Onboard Inertial Measurement Unit (IMU): Each MAV uses its IMU to measure acceleration and angular velocity, which are then used to estimate its own motion.

These data streams are fused using an Extended Kalman Filter (EKF). The EKF is well-suited for this task because it can handle the nonlinearities present in the system's state dynamics and measurements. The paper also introduces a novel system model that simplifies the problem to a 2D plane by considering only the horizontal components of motion and height differences.

A key theoretical contribution is the derivation of system observability conditions. These conditions mathematically define the necessary motion constraints for the MAV formation to ensure accurate and reliable relative positioning. The paper proves that if these conditions are met, the system's positioning accuracy is guaranteed.

Datasets

Not discussed in the paper. The evaluation relies on simulated data generated with parameters derived from sensor datasheets and previous research.

Key Insights and Ideas

  • Observability is Crucial: The paper strongly emphasizes that the system's ability to achieve accurate and reliable relative positioning is directly tied to the MAVs' flight trajectory satisfying specific "observability conditions." This highlights the importance of motion planning for cooperative MAV missions in GNSS-denied scenarios.
  • Ranging and IMU Fusion is Powerful: Combining inter-vehicle distance measurements with IMU data effectively mitigates the drift inherent in IMU-only navigation, leading to significantly improved accuracy and robustness.
  • No Anchor Needed: A major advantage is the elimination of the need for external reference anchors, which broadens the applicability of the method to a wider range of environments and formation sizes.
  • Simplified 2D Model: By leveraging barometric altimeters for height and focusing on horizontal relative positions, the system model is simplified, making the estimation problem more tractable without sacrificing essential information for relative positioning.

Key Results

  • High Accuracy: The proposed method achieves a positioning accuracy of approximately 0.55 m.
  • Significant Improvement: This accuracy is about 3.89 times higher than an existing positioning method ([32]).
  • Robustness to Ranging Errors: Even with ranging errors up to 8 m, the system remains stable and maintains positioning accuracy within approximately 2 m after 100 seconds, demonstrating limited impact of ranging errors on overall performance.
  • Observability Validation: Simulations clearly show that trajectories satisfying the observability conditions lead to significantly smaller and more stable positioning errors compared to trajectories that violate them, which exhibit divergent errors.

Limitations and Future Work

  • Simulation vs. Reality: The authors acknowledge that simulations, while carefully parameterized, cannot fully replicate real-world complexities. Factors like wind, atmospheric variations (affecting barometric altimeters), and airflow from MAV movement can impact practical performance.
  • Future Work: The paper suggests that actual flight verification is the next crucial step to bridge the gap between simulation and real-world performance. Further research could also explore extending the method to formations of more than two MAVs and investigating its performance in more dynamic and cluttered environments.

Summary updated

BibTeX
@misc{cheng2023novel,
  author = {Jian Cheng and Peng Ren and Tingxiang Deng},
  title = {{A Novel Ranging and IMU-Based Method for Relative Positioning of Two-MAV Formation in GNSS-Denied Environments}},
  year = {2023},
  doi = {10.3390/s23094366},
}
Figure