Loading...

Increasing Robustness and Performance of the Ins-Dvl Underwater Navigation by Using Model-Aided Navigation

Karmozdi, Ali | 2021

672 Viewed
  1. Type of Document: Ph.D. Dissertation
  2. Language: Farsi
  3. Document No: 53665 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Salarieh, Hassan; Alasty, Aria; Hashemi, Mojtaba
  7. Abstract:
  8. Dopller velocity log (DVL) is an important underwater navigation sensor. But in the INS-DVL integrated navigation systems, positional error grows unlimitedly over time. Also, the losing bottom lock (LBL) phenomenon is possible in DVL for various reasons such as AUV severe angular changes. So INS-DVL systems are not robust against environmental disturbances resulting in LBL. In this research, mode-aided navigation is used to improve the INS-DVL navigation system. Model-aided navigation is divided into two categories: using kinematic constraints and using kinetic model. Both approaches are used in this research to improve navigation performance and robustness. Initially, the existing kinematic constraints considered in the navigation equations as pseudo-measurement in the Kalman filter. The field test verified successfully improvement in the INS-DVL navigation performance and robustness against the initialization error in the navigation states. Next, the rotational vehicle dynamic motion in the form of a secondary Kalman filter modifies the gyro outputs and enhances the performance of the INS-DVL navigation system. The rotary dynamic model improves the INS-DVL navigation positional error up to 8 times in the presence of gyro bias. The translational vehicle dynamic model is then introduced as a low-cost alternative for DVL that prevents INS instability. It can also make the INS-DVL navigation system robust against DVL outage. Since the model-aided navigation does not depend on expensive additional equipment and does not impose much computational load on the processor, using it to improve the robustness (against initialization error and DVL LBL) and performance of the navigation systems is a good solution
  9. Keywords:
  10. Underwater Navigation ; Model-aided Navigation ; Doppler Velocity Log (DVL) ; Integrated Navigation System ; Inertial Navigation System ; Rotational Dynamic ; Linear Dynamic Analysis

 Digital Object List

 Bookmark

No TOC