The paper proposes a new hybrid approach for quality management of a high-tech production based on a multicolony “Fish School Search” and fuzzy logic. The specific features of quality management of chemical-technological processes necessitate the automation of operational tasks, the manual implementation of which is difficult to implement due to processing large amounts of data and considering many criteria for developing effective solutions. The developed algorithm is based on multicoloniality, which involves parallel work of several colonies of agents, each of which searches for a solution according to a separate criterion, thereby accelerating the process of finding the best result. To verify the effectiveness of the developed algorithm, the results of its testing on a data set are presented, which are sufficiently consistent with the results of manual control by an expert. The conducted simulation experiments showed a sufficiently high accuracy and faster convergence rate compared to the results of a single-population algorithm, which confirms the feasibility of its practical application in quality management of a chemical-technological processes.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Fuzzy Multicolony Bioheuristics for Multicriteria Optimization of Quality Management of Сhemical-Technological Processes

  • Olga V. Bulygina,
  • Andrey M. Sokolov,
  • Margarita Yu. Vorotilova

摘要

The paper proposes a new hybrid approach for quality management of a high-tech production based on a multicolony “Fish School Search” and fuzzy logic. The specific features of quality management of chemical-technological processes necessitate the automation of operational tasks, the manual implementation of which is difficult to implement due to processing large amounts of data and considering many criteria for developing effective solutions. The developed algorithm is based on multicoloniality, which involves parallel work of several colonies of agents, each of which searches for a solution according to a separate criterion, thereby accelerating the process of finding the best result. To verify the effectiveness of the developed algorithm, the results of its testing on a data set are presented, which are sufficiently consistent with the results of manual control by an expert. The conducted simulation experiments showed a sufficiently high accuracy and faster convergence rate compared to the results of a single-population algorithm, which confirms the feasibility of its practical application in quality management of a chemical-technological processes.