中文摘要：

提出了一种基于新的适应度函数的多目标进化算法.该算法通过使用新的适应度函数可以求得分布宽广均匀并且收敛到Pareto前沿的Pareto最优解.首先给出一组分布均匀的向量,然后利用这组向量将种群中的个体进行分类,每个向量作为一类.个体的适应度值是由它所在类的大小及其标量函数值所组成,个体的适应度值越小越好.通过数值试验,该算法与NSGAII,SPEA2和PESAII进行了比较.实验结果表明：该算法对大多数问题可以获得分布更均匀和收敛性更好的解,并且算法运行速度快了很多.

英文摘要：

A novel fitness function-based evolutionary algorithm was presented to solve multi-objective optimization problems.The algorithm could obtain approximate the Pareto frontier and evenly distribute the solutions over the frontier by using the new fitness function.A uniformly distributed set of vectors was firstly given.Then these vectors were used to classify individuals in the population.When each vector was taken as a class,the size of the class was the fitness value of the individuals in the class.Individual＇s fitness value was as small as possible.Compared with NSGAII,SPEA2,and PESAII,results show that the algorithm can achieve better the distribution of solutions and better convergence in most problems and the algorithm can operate faster.