基于扩展最小资源分配网络的热工过程辨识

针对热工过程中表现出的非线性、时变性、大迟延和大惯性等特点,在分析热工过程辨识实际需要的基础上,采用扩展最小资源分配网络,建立了热工过程的非线性在线网络模型.这种模型能较好地解决神经网络序列在线学习问题,其计算量小、计算精度较高.实例计算验证了这种建模方法的有效性和快速性,为热工过程非线性模型的建立提供了一种新的方法.

Identification of Thermal Process Based on Extended Minimal Resource Allocation Network

Aiming at thermal process characteristics of nonlinear,time variance,great delay and big inertia,based on the actual requirement of thermal process identification,an on-line nonlinear network model adopting extended minimal resource allocation network(EMRAN) was proposed.This model can solve the on-line study problem of neural network series,its calculation amount is little and precision is higher.The example calculation proves validity and speediness of the modeling method which provides a new approach to ...