Modeling And Parameter Identification Of GMM-FBG AC Current Sensor Based On J-A Model

The accuracy and reliability of the current detection technology affect the stable operation of the power system.With the rapid economic development of our country and the increasing voltage level and capacity of the power grid,the requirements of the current detection technology in the power system are gradually improved.In recent years,optical current sensor has become a research hotspot of current detection in power system because of its good electrical insulation and fast frequency response.Based on the characteristics of giant magnetostrictive material(GMM)and fiber Bragg grating(FBG),this paper designs a GMM-FBG AC current sensor,builds a GMM-FBG current sensor system and completes the corresponding experimental analysis.The main research work of this paper includes:A lot of domestic and foreign references on GMM and FBG are reviewed,their sensing principles and sensing characteristics are studied,and their respective sensing models are analyzed.A GMM-FBG AC current sensor is designed and the corresponding current sensor model is established.Since the inherent hysteresis nonlinearity of GMM affects the output of the current sensor,an improved J-A model suitable for low frequency is proposed based on the classic J-A hysteresis model.The particle swarm optimization(PSO)algorithm is used to identify and optimize the segmented parameters of the improved J-A model,and the optimal parameter combination of the improved J-A model is obtained.According to the sensing principle of GMM-FBG current sensing system,the GMM-FBG current sensing system experiment platform is built,and on this basis,the sensing experiment is carried out.The effect of bias magnetic field on the sensor characteristics is verified and analyzed by experiments.The improved J-A model is used to simulate and predict the hysteresis characteristics of the sensor and compensate the hysteresis.The validity of the current sensor model and the correctness of the parameter identification result are proved by analyzing the experimental data.