Application of Multi-order Complementary Filtering Technique in Dynamic Attitude Measurement of Forestry Special Robots

Abstract

Abstract-The wide application foreground of forestry special robots makes it become the hotspot of current research. The attitude angle measurement [1] of the special crawler-type robot is one of the key problems. Based on the complementary filtering algorithm, the attitude and angle measuring circuit and data processing algorithm of the special crawler robot are designed. The signal filtering and preprocessing are used to combine the two sensor data. The calculation method of the key parameters in the complementary filtering algorithm is analyzed. The method is applied to the angle measurement of the special crawler robot, and the verification test is carried out. And the experiment result shows that this experiment method is applicable to forestry robots.Keywords-forestry special robots; multi-order complementary filtering; integral drift; attitude solving I. INTRODUCTION Attitude angle measurement is a prerequisite for the operation and control of special crawler-type robots. The accuracy and speed of the attitude angle measurement [2] will directly affect the stability and reliability of the control algorithm of the special crawler robot. With the miniaturization of the inertial measurement components and the improvement of the computing power of the microprocessors, the low-cost inertial measurement unit (IMU) is widely used in posture measurement of forestry special crawler-type robots, and high accuracy attitude measurement is realized by combining the microprocessor data processing algorithms. IMU is mainly composed of low cost MEMS gyroscope and three axis accelerometer. MEMS gyroscope has good autonomy [3], low power consumption, good mechanical and electrical performance advantages of integration. However, the MEMS gyroscope has the characteristic of temperature drift; the measurement error will be accumulated with the accumulation of time, thus affecting the measurement accuracy. The accelerometer will be affected by the vibration of the robot, aliasing the additional amount of vibration interference. So a single sensor measurement is difficult to get accurate attitude angle.