Using Finite-Element Analysis Results and Field-Programmable Gate Arrays to Accelerate Hybrid Powertrain Controller Validation

2015-01-1154

04/14/2015

Event
SAE 2015 World Congress & Exhibition
Authors Abstract
Content
Test and validation of control systems for hybrid vehicle power trains provide a unique set of challenges. Not only does the electronic control unit (ECU), or pair of ECUs, need to smoothly coordinate power flow between two or more power plants, but it also must handle the power electronics' high-speed dynamics due to PWM signals frequently in the 10-20 kHz range.
The trend in testing all-electric and hybrid-electric ECUs has moved toward using field-programmable gate arrays (FPGAs) as the processing node for simulating inverter and electric motor dynamics in real time. Acting as a purpose-built processor colocated with analog and digital input and output, the FPGA makes it possible for real-time simulation loop rates on the order of one microsecond. Combining the temporal fidelity provided by the FPGA with the model fidelity of a machine model based on finite-element analysis yields a hardware-in-the-loop test system that can replicate the high-speed, nonlinear dynamics required to test a power electronics ECU.
Engineers at Subaru were tasked with developing a hybrid electric power train for their first hybrid electric vehicle. The short timetable required innovative test and validation methods that could significantly reduce the typical test time without sacrificing safety and performance. This paper presents the novel approach used and validates the simulation results against physical test data. The new approach yieldeds more comprehensive test capabilities and a significant reduction in test time as compared to traditional methods.
Meta TagsDetails
DOI
https://doi.org/10.4271/2015-01-1154
Pages
7
Citation
Black, B., Morita, T., Minami, Y., and Farnia, D., "Using Finite-Element Analysis Results and Field-Programmable Gate Arrays to Accelerate Hybrid Powertrain Controller Validation," SAE Technical Paper 2015-01-1154, 2015, https://doi.org/10.4271/2015-01-1154.
Additional Details
Publisher
Published
Apr 14, 2015
Product Code
2015-01-1154
Content Type
Technical Paper
Language
English