Parametrical Adjusting of a Hybrid System for the Vehicle Routing Problem with Time Windows

This work employs statistical methods -- variance and linear regression analysis -- to determine the significance degree of parameters of a hybrid system originally proposed for solving the vehicle routing problem with time windows (VRPTW). It is shown how to adjust those parameters to obtain better and more reliable system performance. As a result, it is shown a new best result for solving the instance RC208 of the Solomon's benchmark considering the minimization of the total distance covered for the VRPTW.