In order to transfer multi-objectives problem into single objective problem and consider the decision making process of the uncertainty and fuzziness, it transferred uncertain linguistic information into uncertain linguistic variables, and used uncertain linguistic variables algorithm for computing. Then it defined the complementary judgment matrix, used multi-indexes uncer- ' tainty sorting method to determine the weights of decision-makers, and transferred discrete levels of objective' s attributes to integrated levels and determined the weights of objectives. After that, it made objective' s values normalized and defined a uncertain preference integrated fitness function of multi-objectives problem based on these objective weights, and used particle swarm optimization algorithm to solve the multi-objectives problem. Finally, it used a case to illustrate the algorithm' s feasibility.%为了将彼此冲突的多目标问题转换为单目标问题,并充分考虑专家决策过程的不确定性和思维的模糊性,将不确定语言信息转换为不确定语言变量,利用不确定语言变量运算法则进行计算,通过可能度的定义来建立可能度互补判断矩阵,采用多指标不确定性排序法确定决策者权重,从而将专家对各目标的离散意见转换为综合意见,进而确定各目标权重.通过对各目标值进行规范化处理,综合各目标权重得到决策者不确定性偏好排序的目标综合适应度函数,将多目标问题转换为单目标问题,进而采用粒子群算法对该问题进行求解.最后通过一个算例来说明该算法的实用性和有效性.
展开▼
