The Optimization Design of Fuzzy Controller Based On an Improved Artificial Immune Algorithm

被引:0
|
作者
Yu Mao [1 ]
Ciaen Jie [1 ]
Dou Li-hua [1 ]
Gan Ming-gang [1 ]
机构
[1] Beijing Inst Technol, Sch Informat Sci & Technol, Dept Automat Control, Beijing 100081, Peoples R China
关键词
Artificial immune system; Fuzzy control; Optimization design;
D O I
10.1109/CCDC.2008.4597957
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
An improved artificial immune algorithm is proposed. A large and a small neighborhood are constructed to perform global and local search respectively, so that two-level neighborhood searching mechanism is realized to ensure the search capabilities of the algorithm, and its combination with the control system is presented. Aiming at the fuzzy controller difficulties in the field of intelligent control, it is suggested a design method that employs the improved artificial immune algorithm to adjust the scale factors and membership functions of the fuzzy controller respectively Then, it is designed a fuzzy controller that is composed of two planar fuzzy controllers by using Sugeno fuzzy reasoning method, which successfully controls an inverted pendulum system. The simulation results show that the improved algorithm for single-stage inverted pendulum has an implementation of effective control, and it makes the designed performance of fuzzy controller greatly improved.
引用
收藏
页码:3383 / 3387
页数:5
相关论文
共 8 条
  • [1] CARLOS ACC, 2002, SOLVING MULTIBOJECTE
  • [2] CASTRO D, 2002, CEC 02 P 2002 C, V1, P699
  • [3] CASTRO D, 2000, P GECC 00 LAS VEG US, V7, P36
  • [4] Costa AM, 2002, IEEE C EVOL COMPUTAT, P920, DOI 10.1109/CEC.2002.1007048
  • [5] Learning and optimization using the clonal selection principle
    de Castro, LN
    Von Zuben, FJ
    [J]. IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION, 2002, 6 (03) : 239 - 251
  • [6] DU HF, 2002, 1 INT C MACH LEARN C, V1, P506
  • [7] GOLDBERG DE, P GEN EVCOMP C
  • [8] WIERZCHON S, 2002, TASK Q, V6, P1