Solving the 0/1 knapsack problem with improved pseudo-parallel genetic algorithm

摘要

An improved pseudo-parallel genetic algorithm which comprehensively adopts three different greedy algorithms is proposed to solve the 0/1 Knapsack problem.The initial population is divided into three subgroups,which is respectively pretreated by three greedy algorithms.Information exchange between subgroups adopts the island model to maintain the diversity of population.Simulation study demonstrates that the improved pseudo-parallel genetic algorithm has stronger global search capability and higher immunity against premature convergence than the traditional hybrid genetic algorithm.