摘要
针对多目标库存控制中决策者对目标偏好的不确定性问题,定义了模糊偏好的效用函数,提出了一种基于模糊偏好的多目标粒子群优化算法的求解方法.该方法在改进的双极偏好粒子群优化算法基础上,引入模糊隶属度函数,给出了模糊偏好效用函数的计算方法;利用该函数对得到的非支配解集进行效用评价,以效用评价值作为选取优化方案的依据.针对连续盘点(r,Q)多目标库存控制模型,仿真实验表明:该方法有效地解决了决策者对库存总成本、年平均缺货率和年平均缺货量目标偏好的不确定性问题,为决策者提供了优化的库存控制方案.
Considering the problem of target preference uncertainty of multi-objective decisionmakers, a utility function based on fuzzy preferences is defined and a multi-objective particle swarm optimization algorithm based on fuzzy preferences is proposed. Based on improvement of bipolar preferences particle swarm optimization algorithm, the fuzzy membership function is introduced and the calculation method of fuzzy preference utility function is given. This function is used in utility evaluation for the non dominated solutions which are obtained by the algorithm. The value of utility evaluation can be taken as a basis for selecting optimization scheme. Applied in the multi-objective inventory control model of continuous-review（r, Q） , the method can effectively solve the uncertainty problem of the inventory cost for decision makers, annual average stock rate, and annual average stock target preferences. It can provide an optimal inventory control solution for decision-makers.
出处
《浙江工业大学学报》
CAS
2012年第3期348-351,共4页
Journal of Zhejiang University of Technology
基金
国家自然科学基金资助项目(61070135)
浙江省大学生科技创新活动计划(新苗人才计划)资助项目(2010R403075)
关键词
模糊偏好
多目标优化
效用函数
高斯变异
库存控制
fuzzy preferences
m,dti objective optimization
utility function
gaussian mutation
inventory control
