Combining Adaptive Junction Control with Simultaneous Green-Light-Optimal-Speed-Advisory

Abstract

We introduce the AGLOSA algorithm (pat. pending) which combines an adaptive control algorithm (A) with Green-Light- Optimal-Speed-Advisory (GLOSA) to achieve near-optimal control for a single junction. The central issue which must be resolved to allow this combination is the need for stable signal plans by GLOSA versus the instability caused by adaptive control. Our algorithm exploits Vehicle-to-Infrastructure Communication (V2I) to extend the planning horizon and create sufficiently stable plans using dynamic programming. In a fullyequipped simulation scenario vehicle time-loss is reduced by between 33% and 72% in comparison to competing algorithms.