A Case-Based Reasoning System to Control Traffic at Signalized Intersections

Abstract

Traffic control systems (TCS) contribute to relieve congestion in cities. Although many optimization and intelligent approaches exist to develop TCS, only a few works have investigated Case Based Reasoning (CBR) to control traffic at signalized intersections. Existing works usually state that the case-base is created using experts’ knowledge but do not specify how this knowledge is acquired and how the case-base is built. In this article, we design a CBR system to control traffic at a single signalized intersection. We develop a hybrid methodology to create the case-base using simulation-optimisation, Condensed Nearest Neighbour algorithm (CNN) and a rule-based system. The algorithm is implemented in Python and applied on an intersection simulated using VISSIM, a state-of-the-art traffic simulation software. The performance of the system is assessed and compared to the Longest Queue First with Maximal Weight Matching (LQF-MWM) algorithm. Results show that the implemented system is able to handle different traffic scenarios with competitive performance.