An investigation of inertial sensors parameters on navigation performance

In this master science thesis the method to determine the error budget of the stochastic errors on inertial sensors when they are used in a navigation algorithm is tried to present. The current literature has some tools for this purpose such as Monte-Carlo or Covarince Analysis. However in this statement the analysis will be done with directly by the linear sensor model as different from the others. That is the main object of this work that to calculate the position error which is only caused by the stochastic errors on inertial sensors, the noise term on the other words. To that end the initial focus of the thesis is modelling the sensor by using its data that means system identification. The static position sensor data and simulated white noise data have been used for modelling the sensor by Canocical Correlation Analysis method. The zero input and zero mean data have been used to election of the other deterministic errors on the sensor systems. The preliminary navigation algorithm is applied to calculate the position with its potential error then it is possible to analyze the performance of the inertial sensor. The significance of this study is the determination of the total position error caused from only the noise term on inertial sensor data which detects the performance considerably level.