Loading...

Auv Parameter Estimation Using Sensor Data Fusion Methods Based on Nonlinear Filters

Ghanipoor, Farhad | 2019

486 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 52131 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Alasty, Aria; Salarieh, Hassan
  7. Abstract:
  8. Nowadays, application of AUVs in marine missions is a crucial research field. To reach a fully automated underwater vehicle, it should suitably be designed, controlled, navigated and its fault is detected immediately. An Identified model of AUV is usable for all of these goals. Thus, in this project, parameters of an appropriate dynamic model of AUV is estimated using augmented state space method and TUKF.By investigating possibility of parameter estimation of 6-DOF model and sub models in different cases, a scenario for estimation of cruise, steering and diving sub model parameters of AUV, using output of DVL and Gyroscope is proposed. Furthermore, planar misalignment between DVL and IMU, due to mounting errors, is estimated simultaneously. Estimation is done by both simulated and experimental data. To validate the results, outputs of sensors is compared to model outputs, in two different maneuvers with identification maneuver. Outputs of planar sub models precisely matched to sensor outputs. Moreover, estimated misalignment is evaluated in INS-DVL navigation by effecting it to output of DVL, which decreased end point error to travel distances by 52 percent. In conclusion, the proposed method is suitable for identification of planar sub model and misalignment, using raw data of sensors, done in a simple maneuver
  9. Keywords:
  10. Autonomous Underwater Vehicle ; Nonlinear Filters ; Parameter Estimation ; System Identification ; Misalignment ; Unscented Kalman Filter ; Inertial Measurement Unite (IMU) ; Doppler Velocity Log (DVL)

 Digital Object List

 Bookmark

No TOC