为客观地综合评价合作伙伴,建立了大批量定制环境下合作伙伴的评价指标体系。在此基础上提出了一种基于小波网络的合作伙伴多指标综合评价方法,并结合案例,对大批量定制企业合作伙伴的选择进行了较为客观的评价。最后,给出了该方法与层次分析法和前馈神经网络法的对比。仿真实验表明,该方法不仅能够模拟专家对合作伙伴进行综合评价,避免评价过程中的人为失误,还具有比前馈神经网络法更好的质量和更高的效率。
To both comprehensively and objectively evaluate a partner's quality in mass customization production, a general evaluation system was set up firstly. Secondly, a corresponding comprehensive evaluation method based on wavelet network was put forward in detail for partner selection assessment. A case study for a mass customization production enterprise was discussed by using the suggested evaluation method. Obviously, the application result of the method was comparatively more objective. Finally, the comparisons between the mentioned approach with Ana lyric Hierarchy Process (AHP) and Back--Propagation Neural Network (BPNN) were discussed. It indicated that the method not only could be used for partner selection and avoid some man--made mistakes, but also had better efficiency and quality than BPNN.