The most economical methods for surveying medium and large inland water bodies are based on sonar and echo sounding systems. Today RTK-GPS or total stations are used to position the echo sounding system.
3D orientation is needed as well to position sonar data.
In this thesis the development of an integrated low-cost navigation system based on GPS and MEMS inertial sensors for the determination of position and attitude of the echo sounding system is investigated. A magnetometer is used to estimate the azimuth and is essential to stabilize the whole measurement system. Low-cost inertial sensors exhibit large errors which can be compensated using information on position and velocity as provided by GPS. The data of the individual sensors are combined in a Kalman filter software. New algorithms to include azimuth observations were added to the software.
The Kalman filter relies on the correct quantification of the noise parameters of the MEMS IMU, which were determined in a laboratory test using the Allan Variance method.
A hydrographic survey was carried out in order to assess the algorithms using real data. The assessment shows that the proposed system should be further improved especially with respect to measurements during sharp turns. However, the results also indicate that the desired accuracy of 10 cm (1-sigma) with water depths of 15 m can be achieved using single-frequency GPS, a MEMS IMU and a magnetometer for positioning and orienting the hydrographic sensors.