情報支援によるDVLキャリブレーション

- DVL velocity measurements are critical to AUV navigation accuracy. - DVL is commonly calibrated before mission start while AUV sails on water surface, receiving GNSS signals. - Conventional Kalman filter-based approaches estimate scale factor and misalignment errors. - In GNSS-denied environments, conventional calibration is impossible, forcing use of uncalibrated DVL measurements. - This work proposes information-aided calibration (IAC) to address this limitation.

- Conventional Kalman filter-based calibration requires GNSS signals. - IAC improves accuracy of conventional Kalman filter-based calibration in GNSS-enabled environments. - IAC enables GNSS-free DVL self-calibration, which conventional methods cannot do. - Achieves up to 20% average improvement in GNSS-enabled environments. - Achieves up to 35% improvement in velocity vector estimation during GNSS-free DVL self-calibration.

- Proposes information-aided calibration (IAC). - Two main contributions: improving conventional Kalman filter-based calibration in GNSS-enabled environments, and enabling GNSS-free DVL self-calibration. - Details of the IAC models are not specified in the abstract. - Uses real-world AUV datasets for evaluation.

- Evaluated using real-world AUV datasets. - In GNSS-enabled environments, IAC models achieve up to 20% average improvement. - In GNSS-free DVL self-calibration, up to 35% improvement in velocity vector estimation. - Overall, improves navigation accuracy, reduces navigation drift, and enhances mission reliability.

- The abstract does not discuss limitations or open questions. - Potential discussion points may include applicability to different AUV platforms or environments, but not stated. - From the abstract, no explicit discussion section is mentioned.

- Conventional Kalman filter-based calibration approaches for DVL. - GNSS-aided DVL calibration methods. - DVL self-calibration techniques. - AUV navigation with DVL and GNSS. - Information-aided calibration (IAC) as proposed in this work.