The topic of my master’s degree is the examination, implementation and evaluatation of different signal processing algorithms in embedded environments. I chose this topic because embedded systems, such as telephones, kitchen devices or car electronics help people every day. These embedded systems are simple, responsible and safe, because they work autonomously. To do this, they have to connect to their physical environment throught sensors. The sensor signal processing is an exciting and indispensable duty during the cost efficent planning of modern embedded systems.
The embedded system, which I am working on, is a two-wheel balancing robot based on a mobile inverted pendulum model. The robot is built on a mitmót rapid prototype development platform. It contains a processor module, a display module, a radio communication module, a motor control module and a sensor module which consists of a rate gyro and a two-axis accelerometer.
In my thesis I introduce the components of the controlling circle of the robot. I present the implemented algorithms of the incilantion angle estimation. These algorithms, the complementary-filter and Kalman-algorithm are based on the data of the rate gyro and the accelerometer. I present the theory of the sensorfusion algorithms, the performed measurments, and the reimplemented algorithms based on the results. I describe the integration of the x-IMU orientation measurement unit into the robot, and the realization of the optical identification. Their measured data can be a reference value for the fusion algorithms and they help to compare and evaluate them. After that I introduce the caracteristic of the motor, and the implemented software based on it. I present a more exact way to speed measurement and finally I introduce the robot controlling application