This study aims to propose a temperature error compensation model based on fusion algorithm.
First, the structure and basic working principle of micro electro mechanical system (MEMS) disk solid wave gyroscope are introduced, followed by a temperature error compensation model based on the fusion algorithm of Sparrow Search Algorithm-Variational Mode Decomposition (SSA-VMD) and Random Forest-Glowworm Swarm Optimization-Gated Recurrent Unit (RF-GSO-GRU). The SSA-VMD algorithm is used to decompose the output signals of the MEMS gyroscope under different temperature conditions to obtain different frequency noise signals. The high-frequency noise signals are discarded, and the intermediate frequency mixed noise signals are filtered using the EKF algorithm. The RF-GSO-GRU algorithm is used to establish a temperature error compensation model to compensate for the low-frequency temperature noise signals. Finally, the optimized signals are reconstructed to obtain the compensated output signals of the MEMS gyroscope.
Based on Allan variance method, the factor Angle random walk (N) decreases from 37.526° to 0.471, and the factor Bias instability (B) decreases from 79.2° to 1.226°/h.
This work presents a temperature compensation model based on SSA-VMD and RF-GSO-GRU fusion algorithm and the temperature drift based on the MEMS disk solid wave gyroscope was reduced by proposed method. This fusion algorithm can be widely applied in the field of temperature error compensation for the MEMS gyroscope.
