An improved hybrid backtracking search algorithm based T-S fuzzy model and its implementation to hydroelectric generating units

作者：Yan, Shuangqing; Zhou, Jianzhong^*; Zheng, Yang; Li, Chaoshun

来源：Neurocomputing, 2018, 275: 2066-2079.

DOI：10.1016/j.neucom.2017.10.036

• Fuzzy model identification
• T-S fuzzy model
• Backtracking search optimization algorithm
• Hybrid encoding
• Fuzzy cluster validity index
• Hydroelectric generating units

In order to realize the automatic selection of rule number in fuzzy space partition in the identification of the Takagi-Sugeno (T-S) fuzzy model, a new fuzzy clustering method is proposed in this paper. With the purpose of searching for the optimal fuzzy cluster number and the corresponding cluster centers simultaneously, a novel improved hybrid backtracking search algorithm (IHBSA) has been proposed by introducing the idea of a hybrid encoding scheme as well as a variable valid length chromosome to this optimization algorithm. With a proper cluster validity index taken as the objective function, IHBSA is capable of partitioning the fuzzy space and identifying premise parameters of the T-S fuzzy model without setting a deterministic value of cluster number as a priori. Through the experimental analysis, it is demonstrated that the proposed method possesses higher approximation accuracy with relatively fewer rule number in comparison with traditional ones. Moreover, the T-S fuzzy model is implemented to the practical data of hydroelectric generating units (HGU), and preponderant trajectory matching performance has also been achieved when utilizing the proposed method.

• 出版日期2018-1-31
• 单位华中科技大学