使用改良型人工蜂群演算法於類神經模糊網路

本論文提出了兩個演算法應用於類神經模糊網絡。這兩個演算法包括是改良型人工蜂群演算法和改良型人工蜂群演算法與混沌映射。本論文分為兩個主要部分。在第一部分中，我們提出一個改良型人工蜂群演算法。在此演算法中，我們採用差分進化演算法的突變策略當作原始的人工蜂群演算法的搜索公式來生成新的解並使用貪婪式選擇來決定引領蜂和跟隨蜂蜜蜂的解。此外，所提出的演算法還採用了一個以獎勵為基礎的輪盤式選擇，此輪盤式選擇將所有解分為可行的解和不可行的解，並利用獎勵和懲罰去改變各個解被選中的概率值。在第二部分中，我們提出一個改良型人工蜂群演算法與混沌映射有效的平衡改良型人工蜂群演算法探索和開發的能力。此演算法與第一部分演算法的不同在於引領蜂結合了差分進化的操作和以生物地理學為基礎優化的遷移操作得以產生新的解。另外為了避免改良型人工蜂群演算法陷入局部最優解，此演算法還利用了混沌映射來解決上述的問題。最後，我們將所提出的兩種演算法應用於類神經模糊網路並實施在各種非線性控制系統的問題上。本篇論文的結果得以證實了所提出的演算法之有效性。

關鍵字

神經模糊網路；人工蜂群演算法；差分進化演算法；以生物地理學為基礎的優化；以獎懲為基礎輪盤式選擇；混沌映射；逼近函數；混沌行為；水浴溫度控制系統

並列摘要

This dissertation proposes two algorithms for neural fuzzy networks (NFN) in nonlinear control problem. The two algorithms are including the improved artificial bee colony (IABC) algorithm, and improved artificial bee colony with chaotic map (IABC_CM) algorithm. This dissertation consists of the two major parts. In the first part, the IABC method is proposed for the NFN model. The IABC adopts operator of differential evolution (DE) as the search strategy of artificial bee colony (ABC) to generate new solution and uses greedy selection to decide better solution for employed and onlooker bees. Furthermore, the IABC also uses a reward-based roulette wheel selection will be initially to divide all solutions suitably into feasible and infeasible solutions; thereafter, it divides them based on feasible and infeasible solutions for the implementation of incentives and punishments. In the second part, the IABC_CM is presented to balance the exploration and exploitation of the IABC algorithm effectively. The IABC_CM combines DE operator and migrate operator of BBO to produce new solution for employed bees. In order to avoid IABC method trapped into local optimum, the IABC_CM is utilized chaotic map to solve the above problem. Finally, the proposed two algorithms are applied to implement NFN model in various nonlinear control system problems. The results of this dissertation demonstrate the effectiveness of the proposed algorithms.

並列關鍵字

Neural fuzzy networks ； artificial bee colony algorithm ； differential evolution ； biogeography-based optimization ； reward-based roulette wheel selection ； chaotic map ； piecewise function ； chaotic behaviors ； water bath temperature system.

參考文獻

[1] Z. Michalewicz and M. Schoenauer, “Evolutionary algorithms for constrained parameter optimization problems,” Evolut. Comput., vol. 4, no. 1, pp. 1-32, 1996.

[2] X. Yao and Y. Liu, “A new evolutionary system for evolving artificial neural networks,” IEEE Trans. on Neural Networks, vol. 8, no. 3, pp. 694-713, May 1997.

[3] A. E. Eiben, R. Hinterding, and Z. Michalewicz, “Parameter control in evolutionary algorithms,” IEEE Trans. on Evolut. Comput., vol. 3, no. 2, pp. 124-141, Jul. 1999.

[4] T. P. Runarsson and X. Yao, “Stochastic ranking for constrained evolutionary optimization,” IEEE Trans. on Evolut. Comput., vol. 4, no. 3, pp. 284-294, Sep. 2000.

[5] J. Robinson and Y. Rahmat-Samii, “Particle swarm optimization in electromagnetics,” IEEE Trans. on Antennas and Propagation, vol. 52, no. 2, pp. 397-407, Feb. 2004.

國際替代計量

使用改良型人工蜂群演算法於類神經模糊網路

未授權

主題瀏覽