INS/DVL Fusion with DVL Based Acceleration Measurements

Orzion Levy,Itzik Klein

2023-08-23

Abstract:Autonomous underwater vehicles (AUVs) are increasingly used in many applications such as oceanographic surveys, mapping, and inspection of underwater structures. To successfully complete those tasks, a Doppler velocity log (DVL) and an inertial navigation system (INS) are utilized to determine the AUV navigation solution. In such fusion, DVL velocity measurement is used to update the navigation states. In this paper, we propose calculating the AUV acceleration vector based on past DVL measurements and using it as an additional update to increase the system's accuracy. Simulations and sea experiments were conducted to demonstrate the efficiency of our approach. The results indicate that the proposed method exhibits rapid convergence and significantly improves the overall performance compared to the baseline INS and DVL fusion approach.

Systems and Control

What problem does this paper attempt to address?

### What problems does this paper attempt to solve?

This paper aims to solve the problems of accuracy and robustness encountered by Autonomous Underwater Vehicles (AUVs) during underwater navigation. Specifically, it proposes an improved method of fusing the Inertial Navigation System (INS) with the Doppler Velocity Log (DVL). It calculates the acceleration vector by using past DVL measurement data and introduces it as additional update information into the navigation filter.

#### Background and problem description

1. **Underwater navigation challenges**:
- The underwater environment is complex, and Global Navigation Satellite System (GNSS) signals cannot penetrate water bodies, so other sensors need to be relied on for navigation.
- Commonly used navigation sensors include Inertial Measurement Units (IMUs), magnetometers, pressure sensors, and DVLs, etc.
- Although the IMU can provide a complete navigation solution, its error will accumulate over time, resulting in the drift phenomenon.
- The DVL provides reliable three - dimensional velocity measurements, but in extreme attitudes or terrain changes, some or all of the DVL acoustic beams may be unavailable, affecting the navigation accuracy.

2. **Limitations of existing methods**:
- Existing INS/DVL fusion methods mainly rely on the DVL's velocity measurements to update the navigation state, ignoring the acceleration information.
- In the case of partial or complete DVL failure, existing methods are difficult to maintain high - precision navigation.

#### Proposed solution

The author proposes a new method to improve the accuracy and robustness of the navigation system through the following steps:

1. **DVL - based acceleration estimation**:
- Using the DVL velocity measurement data within a certain period in the past, estimate the acceleration vector of the AUV through mathematical tools such as Taylor series expansion.
- The formula is as follows:
\[
\tilde{a_d^T} = B^{-1} \left[ \sum_{j = 0}^{n - 1} \tilde{v_d^T_j} \quad \sum_{j = 0}^{n - 1} \tilde{v_d^T_j} \Delta_j \right]
\]
where \(\tilde{v_d^T_j}\) is the \(j\) - th DVL velocity measurement, \(\Delta_j=t_j - t_0\), and \(B\) is a matrix composed of time differences.

2. **Acceleration vector as an additional update**:
- Introduce the estimated acceleration vector as additional external measurement information into the Extended Kalman Filter (EKF), combined with the traditional velocity update.
- This approach can directly estimate the accelerometer error, thereby indirectly improving the gyroscope error and overall navigation accuracy.

3. **Experimental verification**:
- Verify the effectiveness of this method through simulation and sea - trial experiments.
- The experimental results show that the new method significantly improves the convergence speed and overall performance of the system in a short time.

#### Conclusion

By introducing DVL - based acceleration measurements, this paper significantly improves the accuracy and robustness of the INS/DVL fusion navigation system. Especially in the case of partial DVL failure, it can still maintain high navigation accuracy. This method provides new ideas and solutions for the development of underwater navigation technology.