Gao, Yuan (Tongji University), sun, kai (Tongji University), Hong, Jinlong (Tongji University), Na, Xiaoxiang (University of Cambridge), Gao, Bingzhao (Tongji University), Chen, Hong (Tongji University)

Learning-Based Predictive Cruise Control for Fuel Consumption Model in Heavy-Duty Trucks

Scheduled for presentation during the Regular Session "Control of heavy vehicles" (FrAT4), Friday, September 27, 2024, 11:30−11:50, Salon 7

2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC), September 24- 27, 2024, Edmonton, Canada

This information is tentative and subject to change. Compiled on October 7, 2024

Abstract

The Predictive Cruise Control (PCC) algorithm is currently being utilized to address the high energy consumption issue in heavy-duty trucks. The premise for achieving significant energy-saving effectiveness with the PCC algorithm lies in accurately obtaining engine fuel consumption model. In this paper, we develop a data-driven fuel consumption model, which uses Gaussian processes regression to learn the prediction error between static fuel consumption map and actual fuel consumption rate. This model possesses the following characteristics: (i) Improved accuracy in predicting fuel consumption rates; (ii) Online parameter updating; (iii) Assisting PCC algorithm in higher energy-saving rates without increasing computational complexity. Through simulation validation, the learning-based model improves fuel consumption rate prediction accuracy by 26.9%. Moreover, the PCC algorithm, which uses this advanced model, has been compared with Adaptive Cruise Control (ACC) and the original PCC, achieving energy-saving improvements of up to 9.6% and 3.1%, respectively.