Abstract
In electric and hybrid electric vehicles, High Voltage DC (HV DC) bus is shared by multiple power inverters, converters, charger and energy storage systems. Most of these components comprise switching power devices and hence inject certain amount of ripple current on the HV DC bus during their normal operation. Therefore, each of these components must comply with a standard HV DC bus ripple requirement without disturbing the operation and stability of other components and impacting life expectancy of HV filter components and HV battery. The traction inverter being the highest power rated component in the system, is the major source of electrical ripple. The magnitude and frequency of the ripple on the HV DC bus due to traction inverter, not only depends on motor speed and torque, but also on the inverter control strategy and parameters. Hence, it is very essential to develop a comprehensive model of the electric drive system with all the HV components, motor with control system and a detailed HV DC bus architecture. The model is used to optimize the filter component size on the inverter side to meet HV DC bus ripple and stability requirements. An electric drive system model is developed in MATLAB for the Chevrolet Bolt Electric Vehicle (EV) platform. The electric drive system model is a detailed one with HV DC bus mechanization, cable impedances, traction inverter and motor with controller software represented as an S-function block. This electric drive model is then connected with other HV components following the HV mechanization and architecture. This enables to carry out the following items: 1. Estimate HV bus resonance by sweeping frequency and ripple by inputting operating points 2. Optimize bulk capacitor size based on worst case operating points 3. Optimize size of inverter components at high switching frequency operation 4. Analyze the impact of fault conditions on the HV bus and other components 5. Analyze HV bus stability and negative impedance phenomenon, if any. The HV DC bus ripple data obtained using the simulation was compared against the vehicle data which showed good correlation. An optimization study was carried out on sizing the bulk capacitor and other filter components. The validation and optimization results are presented in this paper. In the process of making the simulation model more accurate, additional computational capability and memory are required. Hence, it was essential to make a trade-off between added complexities versus accuracy improvement while building the model. Moreover, it was not necessary to model all the HV DC bus components as the ripple contribution from them is very minimal compared to traction inverter (s).