Abstract:To improve the output stability of fiber optic gyroscopes in complex temperature environments, this paper proposes a temperature compensation model based on the Adaptive Spiral Whale Optimization Algorithm (ASWOA) to optimize support vector regression (SVR) parameters. ASWOA introduces a spiral angle adaptive mechanism and a contraction boundary dynamic adjustment strategy into the standard whale optimization algorithm. By collecting zero bias output and temperature data of fiber optic gyroscope in the range of -40 ° C~+60 ° C through experiments, an SVR model with temperature and temperature rate as inputs and zero bias as output is constructed, and the model parameters are globally optimized using ASWOA. The experimental results show that using the ASWOA-SVR model for temperature compensation of fiber optic gyroscopes significantly improves the zero bias stability and compensation accuracy compared to other comparative algorithms, and the model exhibits superior generalization ability for compensation of different samples.