A navigation system estimates the state of motion of a system. The Generic Inertial Navigation System is a multi-sensor data fusion framework that can be used to design a navigation system in MATLAB/Simulink that combines several sensor measurements into a virtual sensor. An inertial measurement unit (IMU) including a 3D accelerometer and a 3D gyroscope is required to predict the motion. Additional sensors can be integrated with minimal development effort.
Wide application range
The Generic INS can be used worldwide, with the exception of the earth poles.
The earth's rotation is also taken into account.
Easy to extend with additional sensors
Thanks to a sigma-point Kalman filter, (non)-linear measurement models can easily be added.
An automatic code generation feature allows the fast creation of embedded MATLAB functions for the fusion of position, velocity, and orientation sensors.
Different sensor sampling rates
Due to the sequential sensor method, sensor data are processed one after the other and only when they have been received.
Transformation to point of interest
A built-in state transformation can be used to transform the state estimation to a desired point of interest.
SRSSUKF
A square-root spherical simplex unscented kalman filter is used.
It is numerically robust and computationally more efficient than other sigma-point Kalman filters.
In addition, some extensions have been added so that quaternions and angles are handled correctly.
To use the MATLAB package and the Simulink library add the lib and packages directory to the MATLAB path. Take a look to the examples.
Robert Damerius and Torsten Jeinsch, A Generic Inertial Navigation System, 2019 International Interdisciplinary PhD Workshop (IIPhDW), Wismar, Germany, May 2019
@INPROCEEDINGS{Damerius201905:GenericINS,
AUTHOR="Robert Damerius and Torsten Jeinsch",
TITLE="A Generic Inertial Navigation System",
BOOKTITLE="2019 International Interdisciplinary PhD Workshop (IIPhDW)",
ADDRESS="Wismar, Germany",
DAYS=14,
MONTH=may,
YEAR=2019
}