Large-scale electromagnetic design optimization of PM machines over a target operating cycle

Abstract

A novel automated design algorithm for applicationbased optimization of permanent magnet (PM) machines is presented in this paper. The proposed algorithm features precise performance evaluation of the potentially heavily saturated machines at high-energy-throughput operating zones using finite element (FE) techniques. First, the energy consumption function associated with the machine's operating cycle is efficiently modeled by a number of representative load points using a kmeans clustering algorithm. Subsequently, a new approach is developed to assess the performance of the machine at each representative load point with proper control to conform to practical operational constraints imposed by voltage and current limits of the motor-drive system. The developed algorithm is applicable to the optimization of any configuration of PM and synchronous reluctance motors over any conceivable operating cycle. Its effectiveness is demonstrated by optimizing the wellestablished reference/benchmark design represented by the 2004 Toyota Prius IPM motor configuration over a compound operating cycle consisting of common US driving schedules. I. INTRODUCTIONDepending on the particular application, the design optimization methods of electric machines [1-7] aim at realizing a set of objectives under certain performance constraints described in the optimization fitness function. Regardless of the type of motor and the objectives/constraints under which the optimization is performed, whether optimizing for constant torque [6] or constant power [7] operations, the automated design techniques evaluate the associated fitness function at the rated torque and base speed. However, in many applications the optimized motor is to operate with variable frequency drives at various load operating points. A case of particular concern pointed out by numerous studies [8][9][10][11] is in traction applications, where the motor is most likely to operate far away from its rated torque and nominal speed. This is whereas the correlation between the main design variables and motor performance metrics, can be significantly affected by the machine's level of ampere loading and magnetic core saturation [12]. This accentuates the challenges in the design of electric motors with intermittent operating cycles where an optimal design is to maintain high performance under various loading conditions.In previous attempts [10,11], a method known as the cyclic representative points was proposed for efficient modeling of

Robust Design of Induction Machines for High-Speed Electric Freight Locomotive Applications

“…The K-means clustering method with the Elbow method's criteria selection was chosen from among the various approaches. Additional representing points are needed for other approaches, such as the Energy Centre of Gravity or the Geometrical Centre of Gravity, which increases the number of FEM runs [37]. Lastly, robust optimization was implemented using the Taguchi method.…”

Large-Scale Design Optimization of PM Machines Over a Target Operating Cycle

“…Moreover, the design optimization can be performed over any conceivable motor operating cycle, while taking into account the practical operational constraints imposed by the supply voltage and/or the motor current limits [21]. Here, the developed method will be used to optimize the Toyota Prius Gen. 2 IPM motor configuration over a combination of common US driving schedules [22]. This specific motor configuration is chosen here for a twofold purpose.…”