节点文献

基于最小二乘支持向量机方法的复杂人机系统操作员功能状态建模与预测

LS-SVM Based Modeling and Forecast of Operator Functional State in a Human-Machine System

【作者】 秦攀攀

【导师】 张建华;

【作者基本信息】 华东理工大学 , 控制科学与工程, 2011, 硕士

【摘要】 对人机系统操作员功能状态(Operator Functional States, OFS)进行准确评估的关键在于建立具有很强预测能力的数学模型。本文基于采集到的一系列操作员电生理信号及性能数据,采用最小二乘支持向量机(Least Squares Support Vector Machine, LS-SVM)方法对OFS建模。通过网格搜索(Grid Search)和10-折交叉验证(10-fold cross-validation)方法对模型参数进行优化,针对标准LS-SVM的不足,采用Suykens提出的LS-SVM的改进算法,建立了具有稀疏性和鲁棒性的LS-SVM模型,并将LS-SVM与基于遗传算法的模糊建模方法进行结果比较。结果表明,LS-SVM方法具有更好的泛化性能,将其用于OFS评估是有效的。根据模型预测结果,便可以调整控制策略,从而设计和实现智能化人机交互系统。

【Abstract】 The core problem of estimating Operator Functional State (OFS) in human-machine systems is construction of an appropriate mathematical model.This paper adopts Least Squares Support Vector Machine (LS-SVM) for OFS modeling based on a series of electrophysiological signals and operator performance data. model parameters are optimized by grid-search and 10-fold cross validation,and we get spare and robust model use modified LS-SVM algorithm proposed by Suykens.The simulation results show that LS-SVM has better generalization performance than GA-Mamdani,and it is efficient and feasible for OFS estimation use LS-SVM. The final model based on the results is used to adjust control strategies, achieving intelligent human-computer interaction.

节点文献中: 

本文链接的文献网络图示:

本文的引文网络