The Research On GMM-FBG AC Current Sensing Characteristic Based On Preisach Model

The key of power system safe, reliable and economic operation is current accuratedetection technology. As its’operating voltage levels and transmission of energy increase,the optical current sensor, such as fiber Bragg grating (FBG) current sensor based on giantmagnetostrictive material (GMM), has become the focus of research and application. Butas sensing core component, the hysteresis nonlinear characteristic of GMM has an effecton the current sensor’ sensing properties such as measurement repeatability reduced,controllable variation. So the modeling compensation has important theoretical andpractical significance.The main research work includes:On the basis of reading lots of domestic and foreign literatures related optical currentsensor, the FBG sensing properties are analysed comprehensively and in depth, includingstrain、temperature sensing properties and cross sensitive characteristic, establishing thesensing model. At the same time current sensing properties and the factors affecting thesensor performance, mainly the hysteresis characteristics, are analyzed for the GMM,established the corresponding strain sensing model.The drive model and structure parameters are designed on GMM-FBG current sensor;According to the GMM hysteresis nonlinear problem, the modeling is established by thePreisach model; On the basis of the classical Preisach model, a new type of Preisachhysteresis model considering frequency is put forward,which is used to describe the GMMhysteresis effect. The model uses similar arc tangent derivative mathematical expressionto describe density function, considering the frequency dependence in parameter setting,which will be able to more accurately predict GMM dynamic hysteresis process.On the basis of theoretical analysis, experimental sensor system and MATLABmodeling compensation control are set up. Sensor hysteresis characteristics are analysedthrough the test and simulation in unbiased and biased magnetic field, the effectiveness ofmodel is improved to predict hysteresis characteristics, and the sensor measurement erroris controled.