A search space analysis for the waste collection vehicle routing problem with time windows

We address a large scale waste collection vehicle routing problem with time windows (VRPTW) complicated by multiple disposal trips and drivers lunch breaks. A fitness search space analysis for several problem instances of the waste collection VRPTW utilizing both statistical and information theoretic measures is presented. We find that mutation landscapes were more easily distinguishable than crossover landscapes. An interesting result was that the information theoretic results seem similar for all the operators indicating the general structure of the landscapes are very similar.