An Isolated Traffic Signal Design using a GA-based Optimization Technique

The Genetic Algorithm (GA) approach is an evolutionary optimization technique, which is developed based on the fundamental theories of natural selection and evolution. The present study focuses on the design of an isolated traffic signal for a two-phase intersection using the conventional method and GA-based optimization technique for unsaturated traffic flow conditions. The methodology of the study includes formulation of an objective function and the constraints, formulation of the constraint violation coefficient, formulation of the modified objective function, formulation of the fitness index (fi), and the GA operations to determine the best green signal timings. The intrinsic nature of Genetic Algorithms in performing elitism, which assures to carry forward the best solution identified in each generation to the next generation. An example problem was solved to demonstrate elitism using binary genetic algorithm with a single variable approach. In the GA operation, parent strings/chromosomes are selected and the crossover is performed along with mutation to form the new offspring’s. Mutation helps in avoiding the convergence of the solution to local optima. In this operation Fitness Index (FI) values of each strings/chromosomes are used as a measure to identify the parent strings to perform GA operations for the next generation. The results of the study indicate that the proposed technique can be used to optimize the signal timings of an isolated traffic signal, this will influence on reducing the delays at the junctions.