SaberRD Design Example: Electric Vehicle Powertrain Simulation with Detailed JMAG Motor Model

Balaji Siva Prasad Emandi

Introduction:

This design example presents techniques to effectively simulate the Powertrain of an Electric Vehicle (EV) at different levels of abstraction, with simulation goals ranging from global efficiency and thermal analysis over hours of driving cycles, down to the nanosecond details of the inverter switching characteristics and losses. Such techniques can be leveraged to optimize motor and inverter controls, to verify power quality (THD and losses), and to simulate faults.


Central to this powertrain design, a high fidelity PMSM motor model produced with the JMAG FEA solver includes spatial harmonics, flux saturation and frequency dependent iron losses. The motor is fed current by a 3-phase Voltage Source Inverter (VSI) controlled by an FOC algorithm implementing Maximum Torque per Amp (MTPA) and Flux Weakening strategies, and using the interceptive sinusoidal PWM method. The dc link voltage (365V) is realized by two Li-ion batteries in parallel, each having 96 cells in series.

Circuit Snapshot:

Figure below is a snapshot of the top level schematic of the EV Powertrain design.

How To Run the Example:

Download the attached "ev_powertrain.zip" and unzip the contents. Open the powertrain1.ai_dsn in SaberRD and run the experiments listed in the Experiment Analyzer.

This EV powertrain design example comes with 4 levels of abstraction, each allowing specific behaviors of the powertrain to be analyzed most effectively. The different abstraction levels are conveniently navigated through by clicking the properties powertrain1 to powertrain4 on the navigation symbol. Alternately, each design can be opened by going to Start Page or File Menu, selecting "Open Design" and the required design from the unzipped folder.

For more details of the design, how to run the design, etc., refer to the README.pdf file in the attachments.

Note:
This design example works only in SaberRD commercial edition 2017.12 Patch1 and above.