Model Predictive Control for Energy-Efficient Maneuvering of Connected Autonomous Vehicles
Southwest Research Institute® (SwRI®), in collaboration with the University of Michigan and Toyota Motor North America, is developing control strategies and technologies aimed at achieving a 20 percent energy efficiency improvement of a 2017 Toyota Prius Prime plug-in hybrid electric vehicle by 2020. The proposed energy efficiency improvement will be achieved without modifications to the stock powertrain hardware and without any compromise in passenger safety. Our test vehicle will be equipped with a Dedicated-Short-Range-Communication (DSRC) radio, enabling connectivity between vehicles (V2V) and infrastructure (V2I). Our control algorithms will leverage information streams via connectivity and use look-ahead information to plan a smarter power-split between the engine and battery to optimize total energy consumption. Utilizing Signal Phasing and Timing (SPaT) information from traffic signals along with realistic route information, our Eco-Approach and Departure (ECO-AND) algorithms will optimize vehicle velocity while ensuring maximum energy efficiency. We are building a traffic simulator based on actual data from Fort Worth, Texas. The traffic simulator will provide realistic route information to the vehicle, and mimic neighboring connected vehicles. By integrating our traffic simulator on a chassis dynamometer, our team will be able to perform vehicle testing in a controlled environment, run quicker iterations on our control algorithms and test the robustness of our developed algorithms by introducing various disturbances in the route. A parallel tech-to-market effort is underway to collaborate with vehicle manufacturers, Tier-1 suppliers and city transit authorities to help bring the technology to market.