混合和声搜索算法优化SVM的订单优先权评价  

Evaluation model of order priority based on HHS-SVM

作　　者：孙莉[1,2] 徐琪[1] 

机构地区：[1]东华大学旭日工商管理学院,上海200051 [2]宁夏大学数学计算机学院,银川750021

出　　处：《计算机工程与应用》2015年第12期241-245,共5页Computer Engineering and Applications

基　　金：上海市浦江人才计划(No.12PJC020)

摘　　要：订单优先权评价是制订生产计划的关键,针对当前订单优先权评价模型不足,提出一种混合和声搜索算法优化支持向量机的订单优先权评价模型(HHS-SVM)。构建订单优先权评价指标体系,采用支持向量机建立订单优先权评价模型,并采用和声搜索算法优化支持向量机参数,在参数寻优过程中,引入了人工鱼群算法的觅食行为,提高了算法跳出局部最优解的能力和收敛速度,采用仿真实验测试模型的性能。结果表明,相对于对比模型,HHS-SVM提高了订单优先权评价精度,是一种可行、有效的订单优先权评价模型。Evaluation model of order priority is a key problem in production plan. This paper proposes an evaluation mod-el of order priority based on Hybrid Harmony Search algorithm and Support Vector Machine to solve the defect of the tra-ditional evaluation model of order priority. The evaluation index system of order priority is established, and then support vector machine is used to establish evaluation model of order priority while the harmony search algorithm is used to opti-mize the parameters of support vector machine. During the search, the artificial fish-swarm prey behavior is introduced in-to the HS algorithm so as to enhance the ability of escaping from local optimal solution and accelerate the converge speed of the algorithm. The performance of model is tested by the simulation experiments. The results show that, compared with other models, the proposed model has improved the evaluation precision of order priority, and it is a feasible, effective evaluation model for order priority.

关 键 词：订单优先权 供应链 支持向量机 和声搜索算法 人工鱼群算法 

分 类 号：F270.7[经济管理—企业管理]