Presently autonomous robotic vehicles are becoming more widespread, especially unmanned aerial vehicles (UAVs). In these mobile devices the navigational computations often rely on an inertial measurement unit (IMU). An IMU used on an aerial vehicle usually includes a 3D accelerometer and a 3D rate gyro. The quality of the sensors and the signal processing architecture fundamentally determines the quality of the navigation.
The devices that are required for the navigational measurements are commercially available, but the selection of low-cost devices suited for university research is very limited. Because of this, the MTA-SZTAKI developed a measurement unit, the Modular Vehicular Navigation System (MVNS). The first part of my thesis is connected to this system. The device designed in SZTAKI was reproduced for BME with some modifications. My task was to adapt the software to these changes, and to insert the new device into the CAN network of a quadrotor under development. A further task was to prepare the system for calibration.
But the device mentioned before is hard to use in small projects, because of its relatively large size and costs, and its modularity isn't required in all applications. Therefore in the second part of my thesis I began to develop a new measurement unit. The goal was to design and build a compact and low-cost unit, with small size and consumption, but accurate enough for the control of aerial vehicles.
The task included the design and construction of the unit and writing of the software running on the microcontroller. The software had to measure, scale and send the data to a central processing unit via serial port and Serial Peripheral Interface (SPI). A further task was to run calibration algorithms increasing the accuracy of the measurements. When the device was finished I've done measurements to determine accuracy and performed the calibration of the new device.