Analysis on Complementary Filter for Inertial Measurement Unit

[[abstract]]本研究運用互補濾波器(Complementary lter) 結合慣性感測元件(Inertial Measurement Unit IMU) 內的加速度計(Accelerometer) 與陀螺儀(Gyroscope) 的狀態量測估算出承載平台的慣性動量及姿態角。利用加速度計之低頻特性以及陀螺儀的高頻特性，透過互補濾波器整合其訊號以達到最佳的估算效果。透過取樣周期與時間常數的方式決定互補濾波器之權重，以此取決出互補濾波器內的加速度計與陀螺儀的信任邊界，設計出不同的權重式互補濾波器架構。此外，還探討在慣性感測元件中使用卡爾曼濾波器(Kalman lter KF) 進行角度估測計算，並分析其特性。卡爾曼濾波器是被廣泛運用的估測器之一，但能否適應目標物運動狀態之變化，往往取決於其內部模型的假設。在實驗部分，則利用加速度計慣性感測元件及陀螺儀慣性感測元件等實驗器材來取得不同運動狀態之慣性量測值，透過互補濾波器的融合運算功能得到低雜訊與高穩定性的姿態角，驗證所發展的互補濾波器再計算慣性姿態角的效果及可行性。由互補濾波器之結果，來探討在各種權重式互補濾波器中，其加速度計與陀螺儀在不同的信任邊界條件下之特性，由這些實驗結果可以得知互補濾波器架構在不同的運動狀態下，能夠估算出精確且穩定的姿態角，可以證實互補濾波器架構能夠降低外在加速度的干擾，得到較佳的姿態角結果。最後，以不同權重值的互補濾波器架構與卡爾曼濾波器作比較。[[abstract]]In the thesis the use of accelerometer and gyroscope for estimating the attitude of an inertial measurement unit (IMU) is studied Using the low frequency characteristic of accelerometer and the high frequency characteristic of gyroscope a complementary filter combines different characteristic data to achieve better estimation From the sample rate and time constant the weights in the complementary filter can be calculated The weights determine the confidence levels on accelerometer and gyroscope In addition we study the angle estimation using the Kalman filter (KF) in inertial sensors In the experiments we acquire the inertial measurement from accelerometer and gyroscope by using NI myRIO and process the data by Matlab It can be seen that accurate and stable attitude angle can be obtained by the complementary filter It is confirmed that the complementary filter architecture can effectively reduce the interference from external accelerometer Finally a comparison between complementary filters with different weights and Kalman filter is conducted The differences of features in the lters are also studie