Inertial Measurement Unit - Data Fusion and Visualization using MATLAB

The present paper is concerned with the development of an algorithm for the processing of data from gyroscopes and accelerometers such that they together form an attitude sensor. The MATLAB programming environment (only MATLAB in the following) has been chosen for the development of this algorithm. The reason has been the ease with which codes can be created and debugged. With the aid of this programming environment it is also possible to use the not very well known but still simple method of visualizing in real time individual signals from primary sensors as well as the algorithm output, i.e. the sensor attitude in space. One chapter of the paper is therefore devoted to the method of real-time visualization in MATLAB. Signals from the three-axis gyroscope are taken as the primary source of information. A disadvantage of gyroscopes is the incessant small unpredictable change of zero position, which is responsible for the error in attitude determination increasing with time. Using signals from the three-axis accelerometer, the presented algorithm tries to minimize this error. Experimental results for the presented algorithm are given in the paper.