The loss characteristics of grain oriented (GO) silicon steel sheet are measured when the excitation contains DC bias and superposition of multiple harmonics. A statistical parameter is extracted from the excess loss under the action of single frequency sinusoidal waveform and DC bias, and then the calculation method of excess loss is improved based on this parameter, so as to accurately predict the loss characteristics of ferromagnetic materials. The consistency between simulation and measurement verifies the effectiveness and accuracy of the method.

iron loss，DC bias， multiple harmonics，prediction model