CAJ | 학술논문

小波神经网络是一种以小波函数为激励函数的神经网络.现有的小波神经网络局限于低维,本文提出一种适合高维输入的小波神经网络建模方法,对几种小波函数与学习算法进行了比较实验,成功地解决了32维输入的大型多辊热连轧机钢板材质量建模问题.
소파신경망락시일충이소파함수위격려함수적신경망락.현유적소파신경망락국한우저유,본문제출일충괄합고유수입적소파신경망락건모방법,대궤충소파함수여학습산법진행료비교실험,성공지해결료32유수입적대형다곤열련알궤강판재질량건모문제.
Wavelet networks now available are usually limited to problems of small dimension input.In this paper, a wavelet-based neural network (WNN) was introduced for handling high dimension input problems. Several different wavelet functions and different algorithms have been tested and compared. Simulation results demonstrate that the B-spline wavelet function and the Levenberg-Marquardt algorithm are effective. The WNN is taken as the production quality model of large-scale hot steel rolling mill and a 32-input modeling problem is successfully solved.