LQR-based perimeter control approach for urban road traffic networks

Abstract

Continuous increase of traffic demands, often more than the network capacity, is the primary reason for traffic congestion. Recently, the existence of macroscopic fundamental diagram in urban road network has been validated by many research works, and then activated the researches for the perimeter control of the network to realize objective of reducing network congestion. In this paper, the centralized state-feedback control design approach for the perimeter control of the network, with the objective of the network balance, is developed, which can realize the balance of network flows and moreover guarantee the maximum network capacity in saturated state of the network. A state-space model of the network perimeter control, with the number of vehicles in links as state variables and perimeter flows as control inputs, is first proposed. Furthermore, LQR approach is applied for the design of perimeter input. Also, the distributed signal control approach of the network under the proposed perimeter control law is proposed, realizing the integration of the perimeter and signal control of the network.