Abstract:Low-cost gyroscope calibration is essential for ensuring the accuracy and reliability of gyroscope measurements. Stationary calibration estimates the deterministic parts of measurement errors. To this end, a common practice is to average the gyroscope readings during a predefined period and estimate the gyroscope bias. Calibration duration plays a crucial role in performance, therefore, longer periods are preferred. However, some applications require quick startup times and calibration is therefore allowed only for a short time. In this work, we focus on reducing low-cost gyroscope calibration time using deep learning methods. We propose a deep-learning framework and explore the possibilities of using multiple real and virtual gyroscopes to improve the calibration performance of single gyroscopes. To train and validate our approach, we recorded a dataset consisting of 169 hours of gyroscope readings, using 24 gyroscopes of two different brands. We also created a virtual dataset consisting of simulated gyroscope readings. The two datasets were used to evaluate our proposed approach. One of our key achievements in this work is reducing gyroscope calibration time by up to 89% using three low-cost gyroscopes.

What problem does this paper attempt to address?

### Problems the paper attempts to solve The paper aims to reduce the calibration time of low - cost gyroscopes by using deep - learning methods. Specifically, the paper focuses on improving the calibration performance of gyroscopes in a short time, so that the navigation system can be deployed and adapted more quickly in applications that require rapid start - up, such as search and rescue operations, drones and robots, etc. ### Background and motivation 1. **Importance of gyroscope calibration**: - Calibration of low - cost gyroscopes is crucial for ensuring the accuracy and reliability of measurements. - Traditional calibration methods usually take a long time, which is unacceptable for some applications that require rapid start - up. 2. **Limitations of existing calibration methods**: - Laboratory calibration methods, although highly accurate, require expensive equipment. - Field calibration methods, although low - cost, have limited accuracy. 3. **Advantages of deep learning**: - Deep learning can process complex data and learn complex patterns, making it a powerful tool for improving the accuracy and reliability of navigation systems. ### Research objectives - **Improve calibration accuracy**: Improve the calibration performance of gyroscopes by deep - learning methods. - **Reduce calibration time**: Achieve rapid calibration so that the navigation system can be deployed more quickly in time - sensitive applications. ### Methods 1. **Data set**: - 169 hours of real - world data were recorded using 24 gyroscopes (of two different brands). - A virtual data set containing simulated gyroscope readings was created. 2. **Neural network architecture**: - A simple convolutional neural network (CNN) was designed to regress the gyroscope bias. - The calibration performance was improved by increasing the number of input channels and the amount of training data. 3. **Experimental setup**: - **Increasing input channels**: Data from multiple gyroscopes were used as input to increase the number of input channels. - **Increasing training data**: Real and virtual data were combined to increase the amount of training data. ### Results - **Significant reduction in calibration time**: By using three low - cost gyroscopes, the calibration time was reduced by up to 89%. - **Improvement in calibration accuracy**: Within 10 - second and 30 - second calibration times, the calibration accuracy of the deep - learning method was significantly better than that of traditional methods. ### Conclusion This study successfully demonstrated how to use deep - learning methods to improve the calibration performance of low - cost gyroscopes in a short time. This provides an effective solution for applications that require rapid start - up.

Rapid Gyroscope Calibration: A Deep Learning Approach

Data-Driven Gyroscope Calibration

A Learning-based approach for bias elimination in low cost gyroscopes

Gyro-aided Camera-Odometer Online Calibration and Localization

In-Field Gyroscope Autocalibration with Iterative Attitude Estimation

Learning to Compensate for the Drift and Error of Gyroscope in Vehicle Localization

Study on Nonlinear Error Calibration of Fiber Optical Gyroscope Scale Factor Based on LSTM

An Efficient Calibration Method for Triaxial Gyroscope

A MEMS IMU Gyroscope Calibration Method Based on Deep Learning

Observability-Aware Intrinsic and Extrinsic Calibration of LiDAR-IMU Systems

LGC-Net: A Lightweight Gyroscope Calibration Network for Efficient Attitude Estimation

Towards Learning-Based Gyrocompassing

MEMS Gyroscope Multi-Feature Calibration Using Machine Learning Technique

A Direct Algorithm for Multi-Gyroscope Infield Calibration

An Efficient Gyro-Aided Optical Flow Estimation in Fast Rotations with Auto-Calibration

On-site scale factor linearity calibration of MEMS triaxial gyroscopes

Gyro-Net: IMU Gyroscopes Random Errors Compensation Method Based on Deep Learning

Denoising IMU Gyroscopes with Deep Learning for Open-Loop Attitude Estimation

A Field Calibration Approach for Triaxial MEMS Gyroscopes Based on Gravity and Rotation Consistency

Research on Calibration Method of MEMS Gyroscope Mounting Error Based on Large-Range Autocollimator

Low-cost MEMS gyroscope performance improvement under unknown disturbances through deep learning-based array