六轴IMU补偿的磁力计动态稳定校准

磁力计是用于确定机器人姿态的常用传感器之一,在使用过程中极易受到周围环境磁场和量测噪声的干扰。传统的椭球拟合能够校准环境磁场的影响,但是不能抑制量测噪声,且不能实时运行,最新的陀螺仪补偿算法弥补了这些缺点,但是由于陀螺仪存在漂移,所以也会使得磁力计数据发生漂移。文中提出一种使用六轴IMU来补偿校准磁力计的方法,首先通过互补滤波使用加速度修正陀螺仪数据,然后使用修正后的陀螺仪数据对磁力计数据进行旋转以对其进行预测,接着使用扩展卡尔曼滤波融合预测值和磁力计量测值,实现磁力计的动态校准。实验表明,相比于传统的椭球拟合算法,文中算法可以降低磁力计数据的噪声波动,二者的噪声波动分别为2μT和0.5μT;相较于最新的陀螺仪补偿算法,由于加速度计具有长期稳定性,可以抑制磁力计数据的漂移现象,磁力计数据的漂移距离由19.56μT降低到了1.67μT。最终实现了一种稳定、高精度的实时磁力计校准。

Six-axis IMU Compensated Magnetometer Dynamic Stable Calibration

Magnetometer is one of the commonly used sensors for determining the attitude of a robot.It is highly susceptible to interference from ambient magnetic fields and measurement noise during use.The traditional ellipsoid fitting can calibrate the influence of the ambient magnetic field,but cannot suppress the measurement noise,and cannot run in real time.The latest gyro compensation algorithm makes up for these shortcomings,but due to the drift of the gyroscope,the magnetometer data will also drift.This paper proposed a method of compensating the calibration magnetometer using a six-axis IMU.First,the gyroscope data was corrected by acceleration correction through complementary filtering,the magnetometer data was then rotated using the corrected gyroscope data to predict it.And then the EKF was used to fuse the predicted value and the magnetic measurement value to realize the dynamic calibration of the magnetometer.Experiments show that compared with the traditional ellipsoid fitting algorithm,the algorithm can reduce the noise fluctuation of magnetometer data.The noise fluctuations of the two are 2μT and 0.5μT respectively.Compared with the latest gyro compensation algorithm,due to the accelerometer has long-term stability and can suppress the drift of magnetometer data.The drift distance of magnetometer data is reduced from 19.56μT to 1.67μT.Finally,a stable,high-precision real-time magnetometer calibration is achieved.