Development and Investigation of a Hybrid Navigation Solution for a Lander Demonstrator

Kurzfassung

In the scope of this work, an extended Kalman filter is designed to enable INS/GPS (Inertial Navigation System/Global Positioning System) based navigation for a lander demonstrator vehicle. The lander demonstrator’s purpose is test and validation of technologies and procedures for lander operations on extraterrestrial bodies such as the moon. Mathematical concepts of attitude representation in the form of quaternions and direction cosine matrices (DCM) are presented. Relevant navigation coordinate frames are introduced and an inertial coordinate frame is chosen as reference for the filter. Basic statistical concepts helpful to understand Kalman filtering are mentioned. The Kalman filter is introduced. Since the navigation task is nonlinear, an extended Kalman filter is derived to produce an adequate navigation solution. Strapdown implementation driven by an inertial measurement unit (IMU) is used as system model. Six individual sensors, namely accelerometer, gyroscope, GPS, barometer, magnetometer and laser range finder, are slated for use by the lander demonstrator vehicle. The sensors’ functional principles are explained and suitable error models are derived. Their error terms are estimated by the filter. This enables correction of measurements and significantly improves navigation accuracy. The filter accomplishes sensor fusion of the incorporated sensors and thus provides a hybrid navigation solution. Execution at 100 Hz qualifies the system as high rate filter. The filter is tested within a high fidelity simulation environment where interactions of controller, actuators, the lander’s physical properties, sensors and the filter itself are considered. Proper navigation performance is shown by means of Monte Carlo analysis. Algorithm enhancements such as sequential measurement update and UD factorization are investigated and implemented. The effects of delayed measurement signals are analyzed. Based on the results of this work, prospects for future work are provided.