Magnetic distortions due to existing appliances and on-board objects with ferromagnetic materials cause serious bias and deviations in motion estimation by inertial measurement sensors with magnetometers. This problem requires a proper sensor fusion to do motion tracking with a minimal angular error. This letter presents a design of a complementary filter that compensates the strong magnetic effects of on-board ferromagnetic materials. Not only the attached permanent magnets may have serious biases on the magnetometer axes but also there is a magnetic distortion due to soft ferromagnetic materials, i.e., steel. After defining the signals of the inertial/magnetic sensors, the process and measurement models are described and a Kalman filter is constructed. The designed filter can be used for motion tracking in environments with magnetic distortions, and in robot actuators with magnetic parts. The performance of the proposed filter is verified under experiment and compared with conventional filters. Finally, we raise a question about whether the attachment of permanent magnets to inertial measurement sensors can serve as a magnetic shield improving the motion estimation.

中文翻译：

---

车载铁磁材料惯性测量单元的运动估计滤波器

现有设备和带有铁磁材料的车载物体引起的磁畸变会导致严重的偏差和带有磁力计的惯性测量传感器在运动估计中的偏差。这个问题需要适当的传感器融合，以最小的角度误差进行运动跟踪。这封信提出了一种互补滤波器的设计，该滤波器可以补偿板载铁磁材料的强磁效应。不仅所连接的永磁体在磁力计轴上可能会产生严重的偏斜，而且由于软铁磁材料（例如钢）会导致磁畸变。在定义惯性/磁传感器的信号之后，描述了过程和测量模型，并构造了卡尔曼滤波器。设计的滤波器可用于在磁失真的环境中进行运动跟踪，以及带有磁性零件的机器人执行器。所提出的滤波器的性能在实验中得到了验证，并与常规滤波器进行了比较。最后，我们提出一个问题，关于将永磁体连接到惯性测量传感器是否可以用作改善运动估计的磁屏蔽。