Physical validation of predictive acceleration control on a parallel hybrid electric vehicle
Date
2022
Authors
White, Samantha M., author
Bradley, Thomas, advisor
Quinn, Jason, committee member
Daily, Jeremy, committee member
Windom, Bret, committee member
Journal Title
Journal ISSN
Volume Title
Abstract
Previous research has been conducted towards the development of predictive control strategies for Hybrid Electric Vehicles (HEVs). These methods have been shown to be effective in reducing fuel consumption in simulation, but no physical validation has been conducted. This is likely due to the fundamental "curses" of dynamic programming mostly the "curse of dimensionality" wherein the run-time needed to generate the optimal solution renders the method unfit as a real-time control. Predictive Acceleration Event (PAE) control combats the run-time issues associated with dynamic programming based control methods by pre-computing the optimal solutions for common Acceleration Events (AEs). This method was physically implemented on a 2019 Toyota Tacoma that was converted into a Parallel-3 (P3) HEV with limited information on the vehicle, including a reduced access to the vehicle's Controller Area Network (CAN) bus. Results from on-track testing indicate a Fuel Economy (FE) improvement in the range of 7% is possible to achieve using PAE control in the real world. To the author's knowledge this is the first time that this type of testing has ever been implemented on a vehicle in the real world.
Description
Rights Access
Subject
controls
fuel economy
predictive acceleration control
electric vehicle
automotive
hybrid