A demodulation parameter optimization method based on improved ant colony optimization support vector regression (IACO-SVR) algorithm is proposed to address the problems of slow demodulation speed and poor anti-interference ability of optical Fiber sensors in extremely low frequency signal demodulation. Firstly, spectral data of optical Fiber sensors under different temperature and pressure conditions were collected through experiments, and the raw data was preprocessed using Kalman filtering algorithm. Secondly, radial basis function (RBF) is used as the kernel function of SVR, and ant colony optimization algorithm (ACO) is introduced to optimize the parameters of SVR to avoid error amplification caused by the sensitivity matrix method for extremely low frequency signals. Finally, a comparative test was conducted between the IACO-SVR algorithm and the Sensitivity Matrix Method (SMM). The experimental results show that in temperature measurement, the average absolute error (MAE) of SMM is 4.99 ℃, and the average absolute percentage error (MAPE) is 14.46%; The MAE of the IACO-SVR algorithm decreased to 1.91 ℃, and the MAPE decreased to 5.61%. In terms of pressure measurement, the MAE of SMM is 2.40MPa, and the MAPE is 16.63%; The MAE and MAPE of the IACO-SVR algorithm were optimized to 0.73 MPa and 4.92%, respectively. The IACO-SVR algorithm has shown higher accuracy and stability, demonstrating the performance advantage of the proposed algorithm.
