Application of Pareto Front to Evaluate Adaptive Traffic Signal Timing for Multiple Objectives

Abstract

This paper examines the effects of policies on coordinated traffic signal control using a multi-objective framework inspired by the Pareto front concept. The Pareto front describes the set of optimal outcomes in a space defined by multiple objectives. This concept is applied to a nine-intersection signalized corridor in a microsimulation study comparing performance from an array of conventional signal control policies that represent a spectrum of options with performance tradeoffs between locally optimal and system optimal control. This is used to identify a Pareto front using delays for coordinated and non-coordinated movements, which offers a frame of reference for comparing the performance of adaptive control algorithms. Two different real-time adaptive control algorithms, a self-organizing algorithm and a schedule-based algorithm, are examined and their performance compared to the Pareto front of conventional controls. The self-organizing algorithm was found to extend the region of feasible performance beyond the capabilities of the conventional methods in different directions relative to the Pareto front.