IMU Data Processing For Inertial Aided Navigation: A Recurrent Neural Network Based Approach

Ming Zhang,Mingming Zhang,Yiming Chen,Mingyang Li

DOI: https://doi.org/10.48550/arXiv.2103.14286

2021-03-26

Abstract:In this work, we propose a novel method for performing inertial aided navigation, by using deep neural networks (DNNs). To date, most DNN inertial navigation methods focus on the task of inertial odometry, by taking gyroscope and accelerometer readings as input and regressing for integrated IMU poses (i.e., position and orientation). While this design has been successfully applied on a number of applications, it is not of theoretical performance guarantee unless patterned motion is involved. This inevitably leads to significantly reduced accuracy and robustness in certain use cases. To solve this problem, we design a framework to compute observable IMU integration terms using DNNs, followed by the numerical pose integration and sensor fusion to achieve the performance gain. Specifically, we perform detailed analysis on the motion terms in IMU kinematic equations, propose a dedicated network design, loss functions, and training strategies for the IMU data processing, and conduct extensive experiments. The results show that our method is generally applicable and outperforms both traditional and DNN methods by wide margins.

Robotics,Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

This paper attempts to address the problem of improving navigation accuracy and robustness in the processing of low-cost Inertial Measurement Unit (IMU) data. Specifically, most existing deep neural network (DNN)-based inertial navigation methods focus primarily on the inertial odometry task, regressing the integrated IMU pose (i.e., position and orientation) by inputting gyroscope and accelerometer readings. However, this design does not guarantee theoretical performance unless specific patterns of motion are involved, which leads to significant degradation in accuracy and robustness in certain application scenarios.

To overcome these issues, the authors propose a new framework that utilizes DNN to compute observable IMU integration terms, followed by numerical pose integration and sensor fusion to achieve performance improvement. Specifically, the authors conduct a detailed analysis of the motion terms in the IMU motion equations, design specialized network structures, loss functions, and training strategies, and conduct extensive experimental validation. The results show that this method outperforms traditional methods and other DNN-based methods in various applications.