摘要
当前具有生物学控制意义的小脑模型只着重于解决小脑内部的控制机理,而没有具体说明输入模块的工作情况。为了使当前的小脑模型更加完整,基于模糊控制理论和认知小脑模型进行分析,在引入模糊集合概念的基础上,通过将小脑模型与模糊理论相结合,提出了一种用于机械臂运动控制的模糊小脑模型。该模型在小脑模型的输入层引入模糊集合的隶属度概念来更加准确地反映客观世界,它不仅能够像仿生小脑模型一样,通过多次学习,对手臂进行轨迹控制,而且具有更高的完整度。首先对构建新模型所需的认知小脑进行了简单介绍,然后对模糊系统进行了相应的研究,并且构建了新的模型,也就是模糊小脑模型。接着研究了新模型所需的各种映射算法,最后利用MATLAB平台进行了相应的仿真实验。实验结果表明,提出的新模型在几次学习之后,也能像认知小脑模型一样准确地对手臂进行控制,而且具有更高的控制精度。
Currently,cerebellar models with biological significance focus only on the control mechanism within the cerebellum rather than the operation of the input module.In order to make the cerebellar model more complete,we analyze the fuzzy control theory and cognitive cerebellar model.At the same time,we introduce the concept of fuzzy set and propose a fuzzy cerebellar model for robotic manipulator control by combining the cerebellar model with the fuzzy theory.The membership degree of fuzzy set has been introduced into the input layer of cerebellar model to reflect the objective world more accurately.It can move the arm along a specific path like a bionic cerebellar model after learning several times with higher integrity.We firstly introduce the cognitive cerebellum needed to construct the newmodel,and then study the fuzzy system and construct a newmodel which is the fuzzy cerebellar model. Then,the various mapping algorithms needed for the newmodel are studied.Finally,the corresponding simulation experiments are carried out on the MATLAB platform.The experiment shows that the newmodel can control the arm exactly like the cognitive cerebellar model after several studies with higher control accuracy.
作者
张少白
诸明倩
ZHANG Shao-bai;ZHU Ming-qian(School of Computer,Nanjing University of Posts and Telecommunications,Nanjing 210003,China)
出处
《计算机技术与发展》
2018年第7期15-20,共6页
Computer Technology and Development
基金
国家自然科学基金(61271334
61373065)
关键词
认知小脑模型
模糊逻辑
映射算法
手臂控制
cognitive cerebellar model
fuzzy logic
mapping algorithm
arm control
