Multiobjective evolutionary path planning via fuzzy tournamentselection

摘要

The paper introduces a new selection algorithm that can be usedfor evolutionary path planning systems. This new selection algorithmcombines fuzzy inference along with tournament selection to selectcandidate paths (CPs) to be parents based on: (1) the Euclidean distancefrom origin to destination, (2) the sum of the changes in the slope of apath, and (3) the average change in the slope of a path. The authorsprovide a detailed description of the fuzzy inference system used in thenew fuzzy tournament selection algorithm (FTSA) as well as some examplesof its usefulness. They use 12 instances of the FTSA to rank apopulation of CPs using the above criteria. Based on its path rankingcapability, they show how the FTSA can obviate the need for thedevelopment of an explicit multiobjective evaluation function. Finally,they use the FTSA to enhance the performance of an existing evolutionarypath planning system called GEPOA