Streamlining Hybrid Vehicle Control Development with an Efficient MATLAB/Simulink Simulation Platform

Abstract

<div class="section abstract"><div class="htmlview paragraph">Automotive hybrid vehicle controls development is an increasingly complex and challenging task. Therefore, to adequately verify and validate the control algorithms prior to its deployment onto real world testing platforms a robust, scalable, low-maintenance simulation platform is most necessary. The currently available test properties pose major challenges in setup, accessibility, maintenance, complexity, and reusability. The aim of this paper is to present a systematic approach of the initial setup, the adaptation to a vehicle program, and the maintenance of a purely MATLAB/Simulink based simulation platform that alleviates the concerns highlighted above. The platform follows the approach of a level 1 virtualization platform for production intent application software components - without the Run-Time Environment (RTE), Basic Software (BSW), and Microcode Abstraction (MCAL) layers. The application layer interacts with a plant model and multiple soft controllers as necessary to form an integrated closed-loop environment. The scale of the application layer to be included within the simulation can be varied based on the nature of the development intent and could be expanded to include additional software components. Additionally, the work involves developing a toolchain to automate the creation of the framework necessary to run the simulation, which significantly reduces the human effort required to adapt the simulation to extended platforms. The benefits include the optimization of the performance of the central processing unit (CPU) running the simulation, while reducing the dependency on the need for expensive licenses and equipment. The results show this process has a quick turnaround time in adapting to different platforms while simulating a large-scale simulation without the need for a high-performance machine.</div></div>