Suppression Method of Gyroscope Drift Applied in Intelligent Devices
Lingling Qi, Mengyao Wang · 2024
Intelligent devices often obtain heading angles through gyroscopes to achieve path planning during motion. This paper focuses on the problem of inaccurate heading angles in intelligent device path systems and proposes an N -point sliding filtering algorithm for calculating heading angles based on gyroscopes. This algorithm utilizes the nonlinear characteristics of N-point static drift errors of gyroscopes to solve the drift problem in heading angles during intelligent device dynamic motions. The algorithm combines $\mathrm{N}-\mathrm{point}$ sliding windows to obtain real-time calibrated heading angles, together with thresholding techniques to distinguish static and dynamic characteristics, improving the reliability of long-term heading angle operations in intelligent devices. In this paper, experiments were carried out by collecting heading angle data of an intelligent lawn mower traveling in a straight 35 m corridor using Glink. The experimental data shows that with this algorithm, the heading angle deviation of the intelligent lawn mower is only 0.14° after reaching the endpoint, increasing the accuracy by 21° compared to traditional lawn mowers. Additionally, the sliding window implementation leads to smoother and more real-time heading outputs compared to N point interval sampling, effectively solving the zero drift problem during dynamic motion, and can be widely applied in path planning and heading control of intelligent devices.