Detection-Driven Adaptive Interference Mitigation-Based Virtual Gyroscope Using Low-Cost Accelerometer and Magnetometer

The increasing demand for micro-electro-mechanical systems inertial measurement unit (MEMS-IMU) has underscored the limitations of conventional hardware gyroscopes, particularly in cost, power consumption, and size, which have restricted their deployment in low-cost systems. This paper presents a detection-driven adaptive interference mitigation (DAIM) virtual gyroscope methodology that utilizes low-cost accelerometers and magnetometers to enable gyroless angular velocity estimation. The research methodology includes preprocessing sensor data, integrating an adaptive filtering parameter strategy based on motion state detection, and dynamic adjustment of fusion weights through interference detection to counteract external disturbances. Angular velocity is calculated by simulating the rotation matrix using vector cross-products, followed by differentiation. Experimental evaluations, including rotary experiments and anti-interference tests conducted in complex scenarios, demonstrate that the proposed method achieves high accuracy across dynamic conditions, resulting in an 18.94% reduction in Root Mean square Error (RMSE) compared to the Tate-Bryan angle method. Furthermore, the method exhibits resilience against external disturbances, such as an additional acceleration of 1g or a magnetic field strength three times that of the Earth’s. A comparative analysis of the applicability of devices with varying costs indicates that this approach reduces hardware dependency and broadens the potential applications for virtual gyroscopes, thereby providing a promising solution for low-cost systems.