摘要
针对间接测量数学模型难于建立的问题 ,提出了基于径向基函数神经网络的非参数测量模型的建立方法。利用对训练样本的聚类结果来确定基函数的中心 ,使得神经网络在较少的训练样本条件下仍可获得较高的建模精度。应用实例验证了本方法的有效性。
WT5BZ]Aiming at the difficulty of the model establishment of indirect measurement, a new method of measurement model establishment based on radial basis function neural networks is presented. In order to obtain a high modeling precision when the size of training sample is small, the center of radial basis function is decided by clustering results of the training sample, then, the effectiveness of this method is proved by application examples.[WT5HZ]
出处
《光学精密工程》
EI
CAS
CSCD
2000年第4期389-393,共5页
Optics and Precision Engineering
基金
国家自然科学基金!资助项目 ( 5 980 5 0 0 7)
关键词
测量模型
径向基函数
神经网络
聚类
检验
measurement model
radial basis function
neural networks
clustering
