All Source Sensor Integration Using an Extended Kalman Filter
Abstract
The global positioning system (GPS) has become an ubiquitous source for navigation in the modern age, especially since the removal of selective availability at the beginning of this century. The utility of the GPS is unmatched, however GPS is not available in all environments. Heavy reliance on GPS for navigation makes the warfighter increasingly vulnerability as modern warfare continues to evolve. This research provides a method for incorporating measurements from a massive variety of sensors to mitigate GPS dependence. The result is the integration of sensor sets that encompass those examined in recent literature as well as some custom navigation devices. A full-state extended Kalman filter is developed and implemented, accommodating the requirements of the varied sensor sets and scenarios. Some 19 types of sensors are used in multiple quantities including inertial measurement units, single cameras and stereo pairs, 2D and 3D laser scanners, altimeters, 3-axis magnetometers, heading sensors, inclinometers, a stop sign sensor, an odometer, a step sensor, a ranging device, a signal of opportunity sensor, global navigation satellite system sensors, an air data computer, and radio frequency identification devices. Simulation data for all sensors was generated to test filter performance. Additionally, real data was collected and processed from an aircraft, ground vehicles, and a pedestrian. Measurement equations are developed to relate sensor measurements to the navigation states. Each sensor measurement is incorporated into the filter using the Kalman filter measurement update equations. Measurement types are segregated based on whether they observe instantaneous or accumulated state information. Accumulated state measurements are incorporated using delayed-state update equations. All other measurements are incorporated using the numerically robust UD update equations. Simulation results show the expected performance of improved navigation state estimation over time w
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 22, 2012
- Accession Number
- ADA562500
Entities
People
- Timothy R. Penn
Organizations
- Air Force Institute of Technology