Effective parameter tuning for genetic algorithm to solve a real world transportation problem

We present and develop a specific and simple method for effective parameter tuning of Genetic Algorithms. This method is based on simple heuristic principle. It measures through specific statistical tests the relevance of the different parameters related to Genetic Algorithms. It measures also how to choose between different available set of parameters values. The implementation of our method is done using C++ and extensive details are shown. We demonstrate our method on a specific real world application of vehicle routing problem. We found that our method helps to reach good performance of Genetic Algorithm.