Research On The Methods For Hysteresis Simulation And Loss Calculation Of The Soft Magnetic Materials

Soft magnetic materials have been widely used in many electrical devices,such as the electric motors,transformers,inductors,but their intrinsic nonlinear hysteresis property has an significant impact on the excitation current waveform,loss and other performances of the power equipment.On the other hand,the losses of soft magnetic materials directly determines the power transmission efficiency of power equipment,and it is also one of the main heat sources in the power equipment.Therefore,the accurate simulation of hysteresis and loss characteristics for the soft magnetic materials is the prerequisite for accurate evaluation of the performance and global optimization design of power equipment.In this paper,the hysteresis modelling and losses calculation methods for the soft magnetic materials are studied.The research achievements and innovations are as follows:1)A hybrid algorithm based on the simulated annealing(SA)and Levenberg Marquardt(L-M)is proposed for parameter extraction of energtic hysteresis model.This method combines the advantages of SA algorithm's strong global search ability and L-M algorithm's fast local convergence speed,and it get rid of the problems of low precision and time-consuming in the existing parameter extraction methods of energetic hysteresis model.Aiming at the problem that the existing static energetic hysteresis model can not accurately simulate the minor hysteresis loop of the soft magnetic materials,the relationship between the model hysteresis loss coefficient and the coercivity,peak magnetic density of the simulated hysteresis loop is analyzed and deduced.Therefore,an improved static energetic hysteresis model is proposed.The simulation and experiment results show that the improved model has high accuracy in simulating minor hysteresis loops.2)To accurately describe the dynamic hysteresis properties for the soft magnetic materials working under the alternating excitations,the analytical expressions for the magnetic field strengthes corresponding to the eddy current loss and excess loss are derived respectively based on the field separation technique and statisticcal theory of losses(STL),with the perspective of energy conservation.The magnetic field strength related to the hysteresis loss is proposed to be solved by the improved static energetic hysteresis model.Consequently,a new analytical and dynamic energetic hysteresis model is constructed,and a loss calculation method for the soft magnetic materials working under DC bias condition based on the dynamic energetic hysteresis model is proposed.Through simulation and experiment,the accuracy of the proposed dynamic hysteresis model and loss calculation method under DC bias is verified.3)A new loss calculation method considering minor biased loops for the soft magnetic materials working under non-sinusoidal excitation is proposed.This method is established according to the independence of the major loop and its internal biased minor loop.With the simplified static Preisach model,a general formula for the hysteresis loss component is derived,and using the functional relationship between the simplified dynamic Preisach model and the excess loss formula developed by STL,the general identification formula for the excess loss statistical parameter is derived.All the parameters of the model can be identified with only a small amount of data measured under the traditional sinusoidal excitation.Finally,the accuracy and practicability of the proposed method are verified by simulations and experiments.4)To solve the problem that the traditional analytical loss models like Steinmetz formula rely too much on experience and need a large number of experimental data for identifying their model parameters,a new broadband analytical loss model for the soft magnetic materials is therefore devised by introducing the fractional derivative theory and modified static energetic hysteresis model.All the parameters of proposed broadband analytical loss model can be extracted by a small amount of measured data,and the issues of time-consuming and non-convergence in the numerical algorithm for computing the high frequency losses are circumvented.Simulations and experimental results verify the accuracy and efficiency of the proposed loss model.