A modeling tool for sensor selection for inertial navigation systems used in underwater vehicles

摘要

Underwater navigation system consists of inertial sensors aided by position, velocity and pressure sensors so that the inherent drift in the inertial measurements are minimized. The choice of the selection of sensor is the tradeoff between the cost and accuracy requirements of the mission. The paper presents the development of mathematical model of sensors used in inertial measurement systems and position estimation using MATLAB. This virtual tool takes into consideration the characteristics of user defined sensors, user mission profile with velocity and heading changes as inputs and outputs the estimated position and distance travelled with time. With the proposed mathematical model, it is possible to select accelerometers and gyroscopes which are required to cater to the required position accuracy which in turn helps the designer to select required aiding sensors for the navigation system. An experiment is carried out with a strap down inertial measurement unit consisting of an accelerometer of bias error 4 mg and gyroscope of random walk error of 4°/√hour for a one hour period in a defined profile and it is found that difference between the model and the experiment results is within 20%. Studies are done using this model for estimating the positional accuracies for tethered and un-tethered underwater vehicles with different sensor accuracies. The need for aiding sensors is also explained.