標題: 模糊邏輯於集體生產規之應用
Application of Fuzzy Logic in Aggregate Production Planning
作者: 伍漢慶
H. C. Wu
許錫美
H. M. Hsu
工業工程與管理學系
關鍵字: 遺傳基因演算法;模糊邏輯控制;隸屬函數;集體生產規劃;Genetic Algorithm; Fuzzy Logic Control; Membership functions; Aggregate Production Planning
公開日期: 1994
摘要: 集體生產規劃(aggregation production planning; APP)是在給定有限的 規劃幅度內,將工廠的人力及設備資源作最有效利用,以滿足預測之顧客 需求。在有限資源共享之前提下,集體生產規劃須決定每一期的生產和僱 用的最佳水準。雖然,目前已有不少數學模式和啟發式的技巧使用於各種 特定的集體生產規劃策略中。但是,這些模式或方法必須依賴精確的參數 估計與函數關係。實際上,決策者常無法精確的估計這些參數估計而給予 確定的函數關係,因此規劃結果無法反應產業實際狀況,致使管理者認為 這些模式或方法無法代表實際上複雜的運作狀況,而對這些數學模式失去 信心。因模糊邏輯不需精確的參數估計與函數關係,因此,本研究擬利用 模糊邏輯於集體生產規劃,並運用基因遺傳演算法(genetic algorithm; GA)來調整修正模糊邏輯控制(Fuzzy Logical Control;FLC)中的隸屬函數 ,並提出修正式的模糊邏輯控制(FLC)模式,以反應生產規劃人員在實際 從事規劃時所採用的決策過程,進而改善傳統集體生產規劃技術的缺點, 以發展一套較能為管理者接受的且具彈性和實用性的集體生產規劃(APP) 模式。 This study presents a method for combining genetic and fuzzy algorithms to learn optimal membership functions for aggregate production planning problems. Generally, the successful application of fuzzy logic controllers depends on a number of parameters, such as membership functions, that are usually subjective creations. It is shown in this study that suitable membership functions may be designed with the aid of genetic algorithms. In order to simplify the computation of the fuzzy logic control, we adopt the trapezoidal fuzzy number as the shape of membership functions. It is shown that the genetic algorithm enable us to generate an optimal set of parameters for these trapezoidal membership functions. The approach presented here is illustrated by using Holt's HMMS paint factory data. Comparison of results with Rinks and Turksen shows that the proposed approach can be used to produce a favourable result.
URI: http://140.113.39.130/cdrfb3/record/nctu/#NT830030027
http://hdl.handle.net/11536/58791
Appears in Collections:Thesis