Using Automated Planning for Traffic Signals Control

Abstract

Solving traffic congestions represents a high priority is-

What you always wanted to know about the deterministic part of the International Planning Competition (IPC) 2014 (but were too afraid to ask)

“…Previous competitions have helped accelerate developments in planning technology, with past competitors being deployed as effective planning components in real-world applications and intelligent systems; see, for instance, Gulić et al . (2016), Matloob and Soutchanski (2016). The wide range of ready-to-use planners and domain models collected for IPC 2014 will drive further developments in the area, not least in the applications of automated planning.…”

“…In contrast to empirical-driven research normally associated with one research group, this event can be seen as a large community-driven scientific experiment, leaving an open legacy of benchmarks, solvers and experimental results. Previous competitions have helped accelerate developments in planning technology, with past competitors being deployed as effective planning components in real-world applications and intelligent systems; see, for instance, Gulić et al (2016), Matloob and Soutchanski (2016). The wide range of ready-to-use planners and domain models collected for IPC 2014 will drive further developments in the area, not least in the applications of automated planning.…”

Design of an Agent-Based Traffic Control System

“…This plan has not been proven to be better than the fixed time systems even though they are capable of selecting and changing their current plan strategically based on the current information given from the networks rather than change based on the time of the day [6]. In some important situations, they can be driven to run in a specific manner.…”

An Innovative Heuristic for Planning-Based Urban Traffic Control

“…Recently, in order to overcome the aforementioned issues of conventional UTC techniques, the application of AI Planning to help in the management of road traffic has been investigated. Works like [9] and [4] have shown the feasibility of applying planning to deal with unexpected circumstances in urban traffic control, by optimising the length of traffic signal phases in the controlled region in order to achieve some specified high level goals. The results of these works also highlighted that the representation of vehicles through the urban network needs to be performed at a macroscopic level -i.e., no explicit representation of each single vehicleto cope with large volumes of traffic.…”

“…The results of these works also highlighted that the representation of vehicles through the urban network needs to be performed at a macroscopic level -i.e., no explicit representation of each single vehicleto cope with large volumes of traffic. The main choice is then between using PDDL and discretising the traffic density (using a sequence of density descriptors) on road sections, as done in [4], or use a numeric representation of traffic density and explicit continuous flow processes by encoding using the PDDL+ [3] language, as done in [9] and, more recently, in [6]. An advantage in using PDDL+ resides in its accuracy, i.e.…”