Combining an evolutionary algorithm with data mining to solve a single-vehicle routing problem

The aim of this work is to present some alternatives to improve the performance of an Evolutionary Algorithm applied to the problem known as the Oil Collecting Vehicle Routing Problem. Some proposals based on the insertion of Local Search and Data Mining modules in a Genetic Algorithm (GA) are presented. Four algorithms were developed: a Genetic Algorithm, a Genetic Algorithm with a Local Search procedure, a Genetic Algorithm including a Data Mining module and a Genetic Algorithm including Local Search and Data Mining. Experimental results demonstrate that the incorporation of Data Mining and Local Search modules in GA can improve the solution quality produced by this method.