广义二型模糊系统的自组织规则生成方法  

Self-organizing rule generation method for a general type-2 fuzzy system

作　　者：范轶博 赵涛[1] 解相朋 FAN Yibo;ZHAO Tao;XIE Xiangpeng(College of Electrical Engineering,Sichuan University,Chengdu 610065,China)

机构地区：[1]四川大学电气工程学院,四川成都610065

出　　处：《智能系统学报》2024年第3期646-652,共7页CAAI Transactions on Intelligent Systems

基　　金：四川省中央引导地方科技发展项目(2021ZYD0016).

摘　　要：针对广义二型模糊系统在复杂情况中因缺乏专家经验而难以构建合适模糊规则的问题,提出了一种基于输入数据的激活强度来生成广义二型模糊集以及模糊规则的方法。通过数据驱动自组织构建广义二型模糊系统模糊规则,并且使用迭代最小二乘法和梯度下降法优化系统前后件参数。最后,分别在无扰动和施加噪声情况下进行了非线性系统的跟踪仿真,实验结果证明了自组织规则生成的广义二型模糊系统的有效性,并能够以较高精度跟踪参考轨迹。To solve the problem of difficult construction of appropriate fuzzy rules in the generalized type-2 fuzzy system due to a lack of expertise in complex situations,a method for generating generalized type-2 fuzzy sets and fuzzy rules based on the activation intensity of input data was proposed.The fuzzy rules of generalized type-2 fuzzy systems were constructed by a data-driven self-organization strategy,and the parameters of the front and rear parts of the system were optimized by the iterative least squares and gradient descent methods.Finally,the tracking simulation of the nonlinear system was conducted in the conditions of no disturbance and noise disturbance.The experimental results revealed that the generalized type-2 fuzzy system generated by the self-organizing rules is effective,and the reference trajectory can be tracked with high accuracy.

关 键 词：自组织学习 模糊控制 广义二型模糊系统 隶属函数 结构学习 误差驱动 增量学习 递归最小二乘法 

分 类 号：TP273.5[自动化与计算机技术—检测技术与自动化装置]