The sensor makes the fiber bragg grating(FBG) adhibited with the giant magnetostrictive material(GMM). The strain change of GMM is transferred to FBG, which caused by the change of its magnetic field from the current coil, then the current will be detected from modulating the FBG change. The paper succeeds in modulating the sensing signal by the way of connecting FBG with matching GMM, and the signal is processed by wavelet analysis technology, which can reduce the error effect from the noise and get the clear picture. Then multisensor data fusion based on radial basis function (RBF) neural network is used in error compensation of the FBG current sensor. The simulation results illustrate that this method is feasible and more effective.