This week, I formule an aerospace/robotic application for the continuous-discrete filters. It consists of the navigation of a 6DOF platform equipped with inertial sensors (accelerometers and rate-gyros) as well as a camera. The scenario contains many fiducial markers placed at known positions. The visual fiducial system is assumed to be available; it provides indirect measures of the visible markers’ relative position w.r.t. the camera. In this application, we have to estimate the platform position, velocity, attitude, as well as the accelerometer and rate-gyro biases.
Material:
- Chapter 8: Navigation Using Fiducial Markers and Inertial Sensors
- Computational Exercise #4: Navigation Using CDUKF and CDEKF
Previous post of this course:
- Week 1: Syllabus + Introduction
- Week 2: Linear Algebra + Linear Systems
- Week 3: Set + Probability + Random Variables
- Week 4: Random Variables + Random Vectors
- Week 5: Random Vectors + Stochastic Processes
- Week 6: Stochastic Processes + Parameter Estimation
- Week 7: Parameter Estimation + Kalman Filter
- Week 8: Kalman Filter and Some Computational Aspects
- Week 9: Extended Kalman Filter
- Week 10: Unscented Kalman Filter
Permalink
Harris ZJ, Paine TM, Whitcomb LL (2018). Preliminary Evaluation of Null-Space Dynamic Process Model Identification with Application to Cooperative Navigation of Underwater Vehicles. IEEE International Conference on Intelligent Robots and Systems. 3453-3459.