中文摘要：

如何将差分演化算法用于离散领域是该领域的一个重要问题．提出一种适应度平均选择的离散差分演化算法，提出的算法中每个个体有均等的机会被选择用于引导算法的进化，这种选择方式有助于克服贪婪选择操作导致的种群多样性下降过快而使算法易陷入局部最优的问题．最后在多维背包问题上的实验结果表明提出的算法具有良好性能．

英文摘要：

How to apply differential evolution to the discrete field is an important problem. A discrete differential evolution with fitness uniform selection scheme is proposed. In the proposed algorithm, each individual has an equal chance of being selected to guide the algorithm evolve. The selection scheme helps avoid the local optimum problem which is led to by the population diversity rapidly decrease of the greedy selection. Simulation results on multidimensional knack problem show that the proposed algorithm has good performance.