一個基於蜂群演算法之新式模糊建模方法

A Novel Fuzzy Modeling Method Based on Bee Colony Algorithm

張耀元(Yao-Yuan Chang)

指導教授 ： 蔡舜宏

國立臺北科技大學/機電學院/自動化科技研究所/碩士(2014年)

https://doi.org/10.6841/NTUT.2014.00654

摘要

本篇論文中，我們提出一個針對T-S模糊模型的新式建模方法，首先利用主成分分析(principal component analysis, PCA)對資料矩陣做線性轉換，以了解資料分佈特性。此外，利用模糊c-平均值(fuzzy c-means, FCM)演算法將資料點做分群。並透過Xie-Beni分群指標找出合適的分群數目以當作線性子系統數目。並利用蜜蜂演算法(bee colony algorithm, BCA)找到優化後的參數值，再結合模糊c-回歸模型演算法(fuzzy c-regression model, FCRM)找出輸入與輸出資料點對這些線性系統的模糊關係，最後再利用權重遞推最小平方法有根據地推算進行修正，經過疊代與更新系統參數得最佳T-S模糊模型參數。經由數個模型進行建模，並比較其他建模方法的結果可驗證所提出之建模方法比其它建模方法更加精確。

關鍵字

T-S模糊模型 ； 模糊分群 ； 蜂群演算法 ； 參數辨識

並列摘要

A novel modeling method for T-S fuzzy model is proposed in this thesis. By utilizing principal component analysis (PCA) the data matrices are linear transformed to the other data distribution matrices. In addition, fuzzy c-means algorithm (FCM) is adopted to classify these data matrices and through Xie-Beni index, the number of the cluster can be determined. Furthermore, by defining the cluster numbers as the rule number, the unknown system can be represented by several linear subsystems. Furthermore, Bee colony algorithm (BCA) and fuzzy c-regression model (FCRM) algorithm are adopted to find the fuzzy relationship between the data points and these linear subsystems. Finally, the weight recursive least squares method is applied to obtain the optimal parameter values of the T-S fuzzy model. Some models are illustrated to demonstrate that our modeling method can provide the better approximation results than some studies.

並列關鍵字

T-S fuzzy model ； Fuzzy clustering ； Bee colony algorithm ； Parameter identification

參考文獻

[1] L. A. Zadeh, “Fuzzy sets,” Information and Control, vol. 8, no. 3, June 1965, pp. 338-353.

[2] J. Q. Chen and L. J. Chen, “An on-line identification algorithm for fuzzy systems,” Fuzzy Sets and Systems, vol. 64, no. 1, May 1994, pp. 63-72.

[3] W. Pedrycz, “Identification algorithm in fuzzy relational systems,” Fuzzy Sets and Systems, vol. 13, no. 2, July 1984, pp. 153-167.

[4] C. W. Xu and Y. Z. Lu, “Fuzzy model identification and self-learning for dynamic systems,” IEEE Trans. Systems, vol. 17, no. 4, Aug.1987, pp. 683-689.

[5] S. K. Sin and J. P. Rui, “Fuzzy system design through fuzzy clustering and optimal pre defuzzification,” IEEE International Conference on Fuzzy Systems, vol. 1, 1993, pp. 190-195.