使用IDG500與ISZ500陀螺儀之方向感測

Abstract

Nowadays, the gyroscopes are used in many mobile systems to sense the orientation change. In sensor networks, the sensors are requested to be small and low-cost. Unfortunately, this kind of gyro usually has the drift problem and the integral heading angle could not be used in application directly. Some calibration methods need to be taken before the gyro output rate being integral to be the heading angle. In our work, we focus on noise filtering, input voltage and temperature calibration, and scalar factor. Besides, because our target application is RSSI finger printing based indoor localization system, which using RSSI finger print to locate person, it is found that the RSSI values are impacted by the human’s direction. Thus, we need to use gyro to get the orientation information even though the gyro output is not that reliable. One of the characters of the indoor moving pattern is that human turning is restricted to right angle in buildings. With this limit, we further proposed a turn detection method that assumes the user can only face four directions. The algorithm output the estimated heading angle and the turning state that determine if it is now turning. With this algorithm, we can estimate the heading angle accurately when we walked along the hallway of our EE building with our daily life paths.