Kan Zhou

Computationally Efficient 3-D Finite-Element-Based Dynamic Thermal Models of Electric Machines

The performance of an electric machine is significantly constrained by temperature. Hence, in order to determine the torque and power capabilities of an electric machine under realtime operating conditions, dynamic knowledge of internal temperatures is required and needs to be estimated in a computationally efficient manner. In this paper, we present a technique for developing computationally efficient thermal models for electric machines that can be used for real-time thermal observers and electrified vehicle powertrain-level simulation and optimization. The technique is based on simulating eigenmodes of the thermal dynamics as determined by 3-D finite-element analysis (FEA). The order of the FE model is then dramatically reduced. The full-order system is decomposed into two parts by using the orthogonality property of the thermal eigenmodes, and only eigenmodes, which are significantly excited, are included in the dynamic model; other eigenmodes are treated as static modes. A large 3-D FEA model can be thus reduced to a small reduced-order model without the necessity of calculating all the eigenmodes. Furthermore, the process of selecting the significantly excited eigenmodes is automatic based on a proposed normalized “extent of excitation” calculation. By using the proposed techniques, the computation time of the model can be dramatically reduced compared with the full-order model while maintaining sufficient accuracy. Experimental results show good agreement between simulation results and measured data.

Finite-Element-Based Computationally Efficient Scalable Electric Machine Model Suitable for Electrified Powertrain Simulation and Optimization

Electric machines are a key component of electric/hybrid electric vehicle (EV/HEV) powertrains. Thus, computationally efficient models for electric machines are essential for powertrain-level design, simulation, and optimization. In this paper, a finite-element-based method for quickly generating torque-speed curves and efficiency maps for electric machines is presented. First, magnetostatic finite-element analysis (FEA) is conducted on a “base” machine design. This analysis produces torque, normalized losses, flux linkage, and the maximum magnetic field intensity in the permanent magnets for a wide range of current magnitudes and phase angles. These values are then scaled based upon changing the size of the machine and the effective number of turns of the machine windings to quickly generate a variety of new machine designs and their corresponding efficiency maps using postprocessing techniques. Results suggest that, by avoiding resolving the FEA for the scaled designs, the proposed techniques can be used to quickly generate efficiency maps, and thus are useful for EV/HEV powertrain-level simulation and optimization.

Computationally-efficient 3D finite-element-based dynamic thermal models of electric machines

Knowledge of the internal temperatures of an electric machine under real-time operating conditions would be extremely useful in order to determine its torque capabilities. This knowledge is also useful for full-scale electric or hybrid electric vehicle (EV/HEV) simulation and optimization. In this paper, we present a technique for developing computationally-efficient thermal models for electric machines that can be used for real-time thermal observers and EV/HEV powertrain-level simulation and optimization. The technique is based upon simulating eigenmodes of the thermal dynamics as determined by 3D finite element analysis. The order of the model is then dramatically reduced in two ways. First, the dynamic system is decomposed into two parts by using the orthogonality property of the eigenvectors. The extent of excitation of each eigenmode is calculated, and only eigenmodes that are significantly excited are included in the dynamic model; other eigenmodes are treated as static modes. Second, only the temperatures in few “hot spots” in various regions of the machine are chosen. By using the proposed model order reduction techniques, the computation time of the model is shown to be reduced by over five orders of magnitude, while maintaining sufficient accuracy. Experimental work also shows a good agreement between simulation results and measured data.

Computationally-efficient finite-element-based thermal models of electric machines

Knowledge of the internal temperatures of an electric machine under real-time operating conditions would be extremely useful in order to determine its torque capabilities. This knowledge is also useful for full-scale electric vehicle simulation and optimization. In this paper we present a technique for developing computationally-efficient thermal models for electric machines that can be used for real-time thermal observers and vehicle-level simulation and optimization. The technique is based upon simulating the eigenmode of the system as determined by finite element analysis. The order of the model is then dramatically reduced in two ways. First, the extent of excitation of each mode is calculated, and only eigenmodes that are significantly excited are included in the dynamic model; other eigenmodes are treated as static modes. Second, only a few “hot spots” in various regions are chosen. The result is a thermal model that can accurately model internal temperatures of the machine while requiring the modeling of only a handful of states. Such a model can be used in vehicle simulations, or for real-time observers in actual vehicles. The computation time of the model presented in this paper is dramatically reduced compared with a typical full-order finite element model.

A stator current locus approach to induction machine parameter estimation

Induction machines are an attractive alternative to permanent magnet machines in transportation applications due to their ruggedness and lower cost of construction. However, high-performance control of induction machines requires accurate knowledge of the machine parameters. While parameters have traditionally been identified using the standard “no-load” and “locked-rotor” tests, these tests are not representative of normal operating conditions, and are not well suited to the commissioning of variable speed drives. This paper presents a new method for induction machine parameter identification using steady-state measurements, which is based on fitting data to the stator current locus for various slip frequencies. Numerical results confirm the methods high accuracy, while experimental results demonstrate its ability to characterize an induction machine over a range of flux levels which include magnetic saturation.

Finite-element-based computationally-efficient electric machine model suitable for use in electrified vehicle powertrain design optimization

Electric machines and their corresponding power electronic drives are key components of electric/hybrid electric vehicle (EV/HEV) powertrains. Thus, computationally-efficient models for electric machines and drives are essential for powertrain-level design, simulation, and optimization. In this paper, a finite-element-based method for quickly generating torque-speed curves and efficiency maps for electric machines and drives is presented. First, magneto-static finite element analysis (FEA) is conducted on a “base” machine design. This analysis produces normalized torque, flux linkage, current, and losses for the operating points of interest. These values are then adjusted based upon changing the size of the machine and the effective number of turns of the machine windings to quickly generate a variety of new machine designs and their corresponding efficiency maps. Results suggest that the proposed techniques can be useful for EV/HEV powertrain design and optimization.