Reduction of the number of rules of high-dimensional hybrid fuzzy system and its hypothesis test of the approximation
WANG Guijun1, DUAN Chenxia2, ZHANG Deli3
1. School of Mathematics Science, Tianjin Normal University, Tianjin 300387, China;
2. Xianyang Road Primary School, Nankai District, Tianjin 300110, China;
3. Information Center, Jilin Provincial Institute of Education, Changchun 130022, China
Abstract：The hybrid fuzzy system is a new system model combining Mamdani fuzzy system and T-S fuzzy system by the control parameters. It retains the excellent characteristics of each system, meanwhile greatly reducing the total number of fuzzy rules. To avoid the rule explosion in high-dimensional mixed fuzzy system under increasing input variables, this article proposes an approximation algorithm for a continuous function based on the layered representation of the hybrid fuzzy system. The total number of rules of high-dimensional layered hybrid fuzzy system can be greatly reduced according to the comparison results. In addition, we simulate the actual output of a three-dimensional hybrid fuzzy system through a practical example, and apply the statistical t-hypothesis test to examine the approximation performance of the three-dimensional mixed fuzzy system.
王贵君, 段晨霞, 张德利. 高维分层混合模糊系统的规则缩减及逼近性假设检验[J]. 浙江大学学报（理学版）, 2017, 44(4): 397-402.
WANG Guijun, DUAN Chenxia, ZHANG Deli. Reduction of the number of rules of high-dimensional hybrid fuzzy system and its hypothesis test of the approximation. Journal of ZheJIang University(Science Edition), 2017, 44(4): 397-402.
[1] RAJU G V S, ZHOU J, KISNER R A. Hierarchical fuzzy control[J]. International J Control,1991,54(5):1201-1216.
[2] WANG L X. Universal approximation by hierarchical fuzzy systems[J]. Fuzzy Set and Systems,1998,93(1):223-230.
[3] LIU P Y, LI H X. Hierarchical T-S fuzzy system and its universal approximation[J].Information Sciences,2005,169(3):279-303.
[4] 杜新宇,张乃尧.二叉树型分层模糊系统的等效性分析[J].清华大学学报:自然科学版,2004,44(7):33-36. DU X Y, ZHANG N Y. Equivalence analysis of binary tree-type hierarchical fuzzy systems[J]. Journal of Tsinghua University:Natural Science,2004,44(7):33-36.
[5] 王贵君,段晨霞.广义分层混合模糊系统及其泛逼近性[J].控制理论与应用,2012,29(5):673-680. WANG G J, DUAN C X. Generalized hierarchical hybrid fuzzy systems and their universal approximation[J]. Control Theory & Applications,2012,29(5):673-680.
[6] 朱晓东,王杰.一种新型分层模糊系统及其逼近性能[J].控制与决策,2013,28(10):1559-1563. ZHU X D, WANG J. A new type of hierarchical fuzzy system and its approximation performance[J]. Control and Decision,2013,28(10):1559-1563.
[7] 王贵君,宋巍巍,韩权杰.基于后件直联型分层的广义混合模糊系统及其积分模逼近[J].控制与决策,2015,30(10):1742-1750. WANG G J, SONG W W, HAN Q J. Generalized hybrid fuzzy system based on consequent direct link type hierarchy and its integral norm approximation[J]. Control and Decision,2015,30(10):1742-1750.
[8] 索春凤,王贵君.最大交互数对非齐次T-S模糊系统的潜在影响[J].山东大学学报:理学版,2015,50(8):14-19. SUO C F, WANG G J. The potential influence of maximum interactive number to non-homogeneous T-S fuzzy system[J]. Journal of Shandong University:Science Edition,2015,50(8):14-19.
2017, Vol. 44 
