Modeling Method Of Power Inductor Oriented To Digital Twin Application

With the development and application of third-generation semiconductor material devices represented by Si C and Ga N,the rapid progress of power electronic device for high-frequency,miniaturization,high-power and flexibility has been promoted.Power inductors is one of the key components in power electronic device,its high-accuracy and high-efficiency modeling methods contribute to the inductor intelligent and refined design,which are important forces to promote the power electronic device development.On the other hand,under the national intelligent manufacturing strategic planning,digital twin technology has become a research hotspot.The study of power inductor models oriented to digital twin applications is a new requirement for inductor design under the trend of intelligent manufacturing.It is also an important development trend of inductance in the future.In this disertation,we propose a power inductor modeling method oriented to digital twins application.Based on a detailed analysis of the flux density distribution features of the power inductor,a magnetic equivalent circuit modeling method is adopted to establish the power inductor model.Based on the distribution features,the inductance model that can fully express the distribution features,has high fidelity and computional efficiency.In addition,it is a practical study for digital twin’s modeling.The main research contents are as follows:(1)Power inductor modeling method for digital twin applications based on distribution featuresWith the finite element method and analytical method,the flux density distribution features of power inductors are studied.Based on it,the lumped parameter method(magnetic equivalent circuit method)is used to establish a power inductor model with distributed features.The model has high accuracy and computational efficiency,and are able to reflect the features of distributed parameters,which make it fulfills the characteristics of digital twin technology for high fidelity,high real-time and future fusion with sensor data.At the same time,under the concept of digital twin growth feature,by establishing a submodel for the local features of the power inductor,the number of features can be further increased and the fidelity of the model can be improved.The modeling method of the power inductor also provides practical method for digital twin modeling application.(2)Topological structure of power inductor model based on distributed featuresRelying on the power inductor modeling method for digital twin applications,the general topology structure of the power inductor twin model is established,which accurately describes the flux density distribution features of power inductors.At the same time,by studying the method of dividing the winding magnetomotive force,the standard winding magnetomotive force branch is established.It makes the shape of the winding and the excitation effect reflected on the inductor model,and the accuracy of the model’s flux density distribution improved,which contributed to increasing the accuracy of inductance and core loss.Furthermore,it provides more accurate information for evaluating the saturation of the inductor.(3)Local model of corners based on distribution featuresAiming at the problems that the flux distribution in core corner is not uniform and varies with the excitation current,which make difficult to accurately calculate the length of the corner magnetic circuit,a local corner model is proposed.By establishing a local magnetic equivalent circuit network for the corner,the equivalent corner branch length can be dynamically changed under different excitation conditions,which overcomes the error caused by the traditional fixed magnetic circuit length of the corner and improves the accuracy.In addition,the local model of corner makes contribution to the accuracy of the power inductor model.(4)Local model of leakage flux outside the core based on distribution featuresDue to the fact that the existing leakage model can not well decribe the morphological features of the winding,a local leakage flux model was constructed based on leakage flux density distribution.This model describes the features of leakage flux outside the core,and has high accuracy for different winding sizes and positions,thereby improving the completeness and fidelity of the inductor model.