基于决策者偏好区域的多目标粒子群算法研究*

多目标优化问题中，决策者往往只对目标空间的某一区域感兴趣，因此需要在这一特定的区域能够得到比较稠密的Pareto解，但传统的方法却找出全部的Pareto前沿，决策效率不高。针对该问题，给出了基于决策者偏好区域的多目标粒子群优化算法。它只求出与决策者偏好区域相关的部分Pareto最优集，从而减少了进化代数，加快收敛速度，有利于决策者进行更有效的决策。算法把解与偏好区域的距离作为影响引导者选择和剪枝策略的一个因素，运用格栅方法实现解在Pareto边界分布的均匀性。仿真结果表明该算法是有效的。

Multi-objective particle swarm optimization algorithm based on DMs preference region

In multi-objective optimization problem,decision makers(DMs) are only interested in a special part of the objective space,so which should has enough solutions.But the traditional approach is to identify all of the Pareto front, so decision-ma-king efficiency is not high. Proposed a multi-objective particle swarm optimization algorithm based on DMs preference region, which only found a preferred and smaller set of Pareto-optimal solutions, insteaded of the entire Pareto frontier, so that reduced the number of iterations and improved the convergencerate, in the end it was beneficial to the decision maker making efficient and reliable decisions. The algorithm took the distance between particle and the DMs preference region as a factor affecting leader selection and pruning strategy, and applied the grid strategy to maintain the diversity of solution. Simulation results show that the proposed algorithm is effective.