Ecological Adaptive Cruise Control for City Buses based on Hybrid Model Predictive Control using PnG and Traffic Light Information

TL;DR

This paper introduces an ecological adaptive cruise control system for city buses that uses hybrid model predictive control and traffic light data to optimize fuel efficiency, vehicle following, and intersection crossing.

Contribution

It presents a novel HMPC-based EACC that integrates PnG and SPaT data for real-time, fuel-efficient bus control in urban environments.

Findings

Significant fuel savings demonstrated in real-world tests.

Controller achieves robust vehicle following and intersection crossing.

Real-time performance confirmed with different prediction horizons.

Abstract

This paper proposes an ecological adaptive cruise control (EACC) concept with the primary goal to minimize the fuel consumption in a city bus with an internal combustion engine (ICE). A hybrid model predictive control (HMPC) is implemented in this work to control both continuous and discrete-time variables. Moreover, a multi-objective optimization problem for EACC is formulated in time-domain as a mixed-integer quadratically constrained quadratic programming (MIQCQP) problem. The proposed HMPC-EACC performs robust vehicle-following while tracking a leading vehicle and plans fuel-efficient acceleration and deceleration maneuvers for the host vehicle. Additionally, it uses the signal phase and timing (SPaT) information to compute a green wave reference speed for the host vehicle to cross the signalized intersections at a green phase. Moreover, the proposed controller performs pulse and glide (PnG) to optimally control the engine ON and OFF states and save additional fuel. Furthermore, the performance of the proposed strategy is evaluated on a real-world driving profile and compared against a baseline controller from the literature. Finally, the influence of different prediction horizons on the fuel savings and computation times are studied. The results reveal significant reduction in fuel consumption with HMPC-EACC and demonstrate that the proposed controller is real-time capable.