Novel Optimization Algorithms for Oversaturated Traffic Network Coordination

Sponsor:  Michigan Tech Transportation Institute (MTTI)

PI:  Ossama Abdelkhalik

The objective of this project is to solve the Oversaturated Traffic Network Signal Coordination Planning (OTNSCP) optimization problem, through a new problem formulation and using the recent HGGA optimization method. The long term goal is to developed software tools that model and optimize traffic networks both off-line and online, taking into consideration the network uncertainties and travelers’ behavior.

The proposed problem formulation handles the OTNSCP optimization problem as a grouping problem, rather than the standard optimization for the individual signals green times. In the off-line optimization of traffic networks using the standard formulation, the number of variables is the number of design green times. In the proposed formulation, it is assumed that some of the traffic signals in the network have the same green time value; in other words a subset of the traffic network signals has the same green time value for all signals in the subset. The network may have several subsets. The number of signals in each subset is a variable to be optimized.

The nature of the OTNSCP problem is that signals should be coordinated (using green times) to maximize the overall network throughput. So, instead of formulating the problem to optimize the individual signals’ green times, it would be more natural and efficient if it is formulated as a distribution of green times over the network of signals, collectively, as a group to maximize the objective function.

Ossama Abdelkhalik