Development of GNSS/IMU UAV-borne scalar gravimetry system

• Autor(a) principal: Ostapchuk, Dmytro
• Data de Publicação: 2023
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10773/41076
• topic: UAV-borne gravimetry; Strapdown IMU; GNSS; Kalman filter

Airborne gravimetry is the determination of the Earth’s gravity field, using aircraft as a mobile measurement platform. Nowadays, airborne measurements cover an important intermediate area between terrestrial and satellite gravity methods in terms of both resolution and coverage. However, the aircraft rental cost is high, and strict rules and regulations exist for acquiring permission to conduct a flight mission in most countries. A possible alternative investigated in this thesis is based on the use of an Unmanned Aerial Vehicle (UAV). Its advantages include high manoeuvrability, operational flexibility and no airport support. The development of a compact size low-cost strapdown scalar gravimetry system for hexacopter UAV is attempted and its performance is evaluated. This gravimetry system is based on a combination of one Global Navigation Satellite System (GNSS) receiver and a (strapdown) Inertial Measurement Unit (IMU) that are the two components of tactical-grade INS Quanta. IMU observations are integrated with respect to time and the gravity disturbance is obtained from the comparison of IMU-predicted position and velocity with GNSSobserved position and velocity, i.e. inertial navigation approach was adopted. Raw GNSS observations are processed into velocity and position estimates using Qinertia® software. All observations, IMU and GNSS data, are processed within a single loosely coupled Extended Kalman Filter (EKF) that includes the system state for gravity. For the implementation of the filter, the random constant process was adopted to model the errors in the gyros and accelerometers. Whereas, the gravity disturbance was modelled as a simple random walk process. In this way, the total state vector was composed of the position (3 states), the velocity (3 states), the attitude (3 states), the accelerometer and gyro biases (6 states), and one state for the gravity disturbance. In the scope of this thesis the UAV-borne gravimetry processing tool called gravito was developed. Evaluation of the system was primarily based on collecting the statistics of innovation sequences and comparing EKF’s estimates of position, velocity and attitude with those obtained by a commercial software Qinertia®. Based on the real campaign data, it was concluded that the internal gyroscope noise of IMU Quanta does not allow the estimation of the heading, as well as the roll and pitch angles, with sufficient accuracy for scalar gravimetry. Also increased dynamics of UAV (in comparison with manned aircraft) add additional errors to the gravity disturbance estimation by making invalid assumptions of linearity for the dynamic model of the system and through the scale factor errors of IMU. In this way, the current low-cost system with the processing method implemented is not suitable for scalar gravimetry mainly due to the noise in the IMU Quanta gyroscopes measurements.