Self-stabilizing platform : ZTäBiLAjZöR

Abstract: The use of accelerometers and gyroscopes has greatly increased in recent years, different applications of these are seen in everything from smartphones to mo-torcycles. The stabilizing mechanisms used for cameras may however be more well-known. Irrespective of purpose, motorized stabilizing mechanisms com-monly utilize an inertial measurement unit to determine real-time position and movement. This position is then analysed and an action is chosen. To improve performance, control science is often applied. The purpose of this report is to analyze how a common PID controller will react to variations in load. This report will describe the construction and development of a self-stabilizing platform, controlled by two PID controllers, based on the Arduino platform. Theory behind the included components is presented and the methods of modeling, construction and testing is discussed. An IMU consisting of three accelerometers and three gyroscopes is used to measure the tilt of the platform, a Kalman filter is then used to suppress the signal noise to reasonable levels. This was successfully achieved using Open- source code from the Arduino community. Two DC-motors adjust the angle of the platform, one for each axis, with a PID controller each. The system was modelled for each motor to determine the PID controller parameters that would satisfy the requirements set regard-ing speed and precision. A fast system was prioritized to assure small angles and reduce the torque on the motor shafts. This resulted in a system with roughly 0.1 seconds of rise time, 1.3 seconds settling time and 18% overshoot at a 10° step. By conducting step response tests using the demonstrator, the research ques-tion could be answered. It was found that a PID controller tuned for a specific system will have a performance drop but not become unstable when load is varied, up to the level of load that causes too large amounts of torque for the motors to counter.