查詢結果分析
來源資料
頁籤選單縮合
題名 | 機率結構如何改善模糊推論系統=Improvements for Fuzzy Inference Systems due to Probabilistic Structures |
---|---|
作者 | 李穎; 蘇伯達; Li, Ying; Su, Po-ta; |
期刊 | 模糊系統學刊 |
出版日期 | 19990600 |
卷期 | 5:1 1999.06[民88.06] |
頁次 | 頁11-21 |
分類號 | 319.9 |
語文 | chi |
關鍵詞 | 模糊集; 模糊推論系統; 機率; Fuzzy sets; Fuzzy inference systems; Probability; |
中文摘要 | 模糊推論系統用途頗廣,可以實現模糊控制、系統鑑別、訊號處理等功能。採用 正規化激發強度,乘積模糊化,乘積推論,以及相加組合之模糊系統可視為具機率結構。以 前的研究側重理論探討,實際應用中採用機率結構有何優點並未進一步討論。本文中我們首 先回顧以機率結構定義模糊集的觀念,再舉例說明採用機率結構對模糊推論系統之影響。機 率結構使模糊推論系統以較合理的方式進行插值,可使模糊系統輸入輸出函數與語言描述配 合較佳,可調參數對輸入輸出函數之影響也較清楚。 |
英文摘要 | Fuzzy inference systems have a wide range of applications, including control, system identification, and signal processing. Fuzzy inference systems with normalized firing strengths, product fuzzification, product inference and sum composition are said to have the probabilistic structure. Past research in this area emphasizes more on theory, there are few discussions in the advantages of probabilistic structure in real applications. In this article, we first review the definition of filzzy sets based on probabilistic structures, then give examples to illustrate how probabilistic structures affect fuzzy inference systems. Probabilistic structures enable fuzzy inference systems to interpolate in a more reasonable manner, enhance the agreement between the input output functions and the linguistic descriptions, and clarify the effect of the adjustable parameters on the input output functions. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。