Group-based approach to adaptive traffic-signal control

Abstract

A group-based adaptive traffic-control method for isolated signalized junctions is developed in which an hierarchical structure is created, consisting of tactical and local levels of signal timing optimization. This control method optimizes signal timings in adaptive traffic-control systems, and takes full advantage of flexible new technologies to incorporate the most up-to-date traffic information, as collected in real time. The cycle structure’s definitions, combinations and sequencing of stages are generated automatically through a procedure for optimizing signal-timing plans in response to online data from traffic detectors. This new method allows a wider search space for more efficient signal-control systems to improve junction performance, and thus minimize delays and maximize capacity in real time.

The contributions developed in this thesis include efficient procedures for estimating essential inputs for adaptive control and for applying detector data to measure turning proportions, prevailing arrival patterns and queue-length distributions. Other contributions involve the development of predictive procedures for estimating lane-based delay and for predicting future lane-based delay on the basis of a rolling-horizon approach. A multi-resolution strategy is proposed for updating the elements of signal plans cycle-by-cycle, and for adjusting the current green signal timing second-by-second. The proposed adaptive traffic-control method comprises six parts: a lane-to-lane turning-flow estimation method, a lane-based queue-length estimation method, a lane-based polygonal delay-estimation formula using a rolling-horizon approach, a lane-based formula for derivatives of delay, a group-based proactive global-optimization procedure and a group-based reactive local signal-control policy.

Methods for analyzing lane-based queue lengths and lane-to-lane turning flows are implemented, as these input data are essential for deriving group-based adaptive controls from incomplete raw detector data collected in real time. The aggregate queuing delay in each lane at the end of each cycle is estimated through a polygonal delay formula (as introduced in HCM (2010)) for the proactive global-optimization of signal timing at the tactical level of control.

The group-based variables and parameters for the proactive global-optimization method utilize lane-based predictive traffic-flow information such as arrival and discharge rates, expressed as the slopes of polygonal delay formulas. Therefore, there is a high degree of flexibility in the tactical identification of the optimal signal plan, in response to the real-time predicted traffic information, the objective function of the polygonal delay formula and the direct differential equations for adaptive group-based variables.

The reactive local signal-control policy, which is formed on the basis of the max-pressure strategy, is developed to locally adjust the current green signal time and to carry out delicate demand fluctuations second-by-second at the fine-resolution level. Therefore, the most appropriate cycle-structure for the tactical level of control is identified through a group-based global-optimization procedure that uses the latest available information.

The results of the computer simulations conducted for this thesis show that the integrated group-based adaptive traffic-signal-control logic outperforms the other methods over a wide range of traffic conditions, from free-flowing traffic to extreme congestion. Moreover, the performances from all of the proposed models are much better than those for the existing fixed-signal plan or the actuated signal-control on asymmetric traffic conditions.