Model-based green time optimisation

This paper presents a green time optimization model at intersections, which uses the expected queue length and delay as its optimization objective. In the process of solving, the generalized reduced gradient algorithm for optimizing nonlinear problems imbedded in Microsoft Excel Solver© was used. By using this method, the optimal value of the phase/signal group green times of the intersection was obtained based on the minimum total delays. A hypothetical example of a three-phase signalized intersection is presented. The result shows that the variation of arrival rate has an obvious effect on allocating the green time, which illustrates potential enhancements in system operation efficiency and fairness under various traffic demands.