• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

计算机工程与科学

• 论文 • 上一篇    下一篇

基于智能算法的技术等级评估方法研究

唐湘滟1,蔡宽麒2,程杰仁1,3,刘博艺1,4   

  1. (1.海南大学信息科学技术学院,海南 海口 570228;2.海南大学机电工程学院,海南 海口 570228;
    3.海南大学南海海洋资源利用国家重点实验室,海南 海口 570228;4.中国科学院大学,北京 100000)
  • 收稿日期:2017-07-14 修回日期:2017-09-18 出版日期:2018-01-25 发布日期:2018-01-25
  • 基金资助:

    国家自然科学基金(61363071,61762033);海南省自然科学基金(617048);海南大学博士启动基金(kyqd1328);海南大学青年基金(qnjj14444);海南大学研究生实践创新项目;南海海洋资源利用国家重点实验室资助

Dicing technical level evaluation
based on intelligent algorithm

TANG Xiang-yan1,CAI Kuai-qi2,CHENG Jie-ren1,3,LIU Bo-yi1,4   

  1. (1.College of Information Science & Technology,Hainan University,Haikou 570228;
    2.College of Mechanical and Electrical Engineering,Hainan University,Haikou 570228;
    3.State Key Laboratory of Marine Resource Utilization in South China Sea,Hainan University,Haikou 570228;
    4.University of Chinese Academy of Sciences,Beijing 100000,China)
  • Received:2017-07-14 Revised:2017-09-18 Online:2018-01-25 Published:2018-01-25

摘要:

胶工的割胶技术水平是影响橡胶产量的关键因素,技术一般的胶工要比技术娴熟的胶工少产20%~30%的橡胶,为此提出了基于智能算法的技术等级评估方法,设计并实现了割胶技术智能辅助学习仪。给出了智能辅助割胶技术学习仪的体系结构并构建了基于智能算法的技术等级评估指标体系,该指标体系利用德尔菲法采集10个评价指标数据,通过灰色关联度分析方法筛选和确定指标,利用熵权法对各项指标赋权,求得割胶水平量化函数。然后,基于逆向云发生器,将割胶水平量化得分转化为割胶水平的定性评价,进行技术等级的分类。最后,利用k-means聚类法确定不同割胶水平的中心点,并利用最小欧氏距离识别胶工割胶水平技术等级。实验结果表明,本文提出的方法评估准确率高,达到90%以上,同时实时性强,评价实时更新时间小于3 s,填补了智能辅助学习割胶技术空白,有利于胶工更快提高割胶水平、增加橡胶产量,具有较好的推广和应用价值。

关键词: 割胶技术, 灰色关联, 熵权法, 等级评估, 云模型

Abstract:

The technical level of rubber working is the key affecting factor of rubber production. Less skilled laborers harvest about 20%~30% less rubber than those technologically skilled ones. For this reason, we put forward a technical level evaluation method based on the intelligent algorithm. We also design a tapping technology intelligent auxiliary learning instrument. Firstly, we explain the architecture of the intelligent tapping technology auxiliary learning instrument, and construct an evaluation index system of technical level based on the intelligent algorithm. The index system uses the Delphi method to collect 10 evaluation index data, filter and determine the index by the gray relational analysis method, and use the entropy weight method to empower the index, thus obtaining the dosing level quantization function. Then, based on the reverse cloud generator, the tapping horizontal quantization scores are transformed into the qualitative evaluation of the tapping level, and the technical grade is classified. Finally, the k-means clustering method is used to determine the center points of different tapping levels, and the minimum level of European distance is used to identify the technical level of glue peeling. Experimental results show that the proposed method has an over 90% accuracy rate and satisfactory real-time performance. The real-time updating time is less than 3 seconds, which fills the gap in the intelligent auxiliary learning tapping technology, which is beneficial for glue workers to improve the tapping level and increase rubber production, thus having a good promotion and application value.


 
 

 

Key words: tapping technology;gray correlation;entropy weight method, technique grade, cloud model