摘要
通过对现有网上谈判模型研究发现,原有求解方法因多采用线性效用函数假设而存在自身无法克服的局限性。提出了基于BPNN-GA算法的双边多属性谈判求解方法。为求解属性间具有复杂函数依赖关系、效用函数难以显式表达,以及具有非线性效用函数映射关系的双边多属性谈判提供了一种有效的求解方法。最后通过谈判实例验证了该方法的有效性。
Through the study of on-line negotiation model,it was found that the original solving method had its own limitations due to the assumption of linear utility function.This paper proposed solution method of bilateral multi-attribute negotiations based on BPNN-GA algorithm.It is an effective solution method for bilateral multi-attribute negotiations with the complex function of inter-dependencies among the attributes,and a non-linear mapping which is difficult to express explicitly in the utility function.Finally the effectiveness and superiority of the method was demonstrated by an example.
出处
《武汉理工大学学报(信息与管理工程版)》
CAS
2011年第2期320-324,共5页
Journal of Wuhan University of Technology:Information & Management Engineering
基金
国家自然科学基金资助项目(70872010)
中央高校基本科研业务费专项基金资助项目(FRF-BR-09-011A)
黑龙江省自然科学基金资助项目(G201021)
关键词
双边多属性谈判
BP神经网络
遗传算法
多属性依赖
bilateral multi-attribute negotiations
BP neural network
GA
multi-attribute dependence