论文部分内容阅读
前馈型神经网应用于非线性系统辨识的一个问题是确定系统阶次。采用前馈神经网进行非线性系统定阶与神经网的推广性问题密切相关。ＯＬＳ算法是构筑径向基神经网的一种学习算法，但是采用ＯＬＳ算法构筑神经网存在推广性问题。ＲＯＬＳ算法将ＯＬＳ算法与正则化（ｒｅｇｕｌａｒｉｚａｔｉｏｎ）方法相结合，以提高算法的推广能力。本文将基于径向基网的ＲＯＬＳ算法应用于非线性系统定阶。本文对提出的方法进行了仿真研究，结果验证了方法的有效性。
One of the problems that feedforward neural networks apply to nonlinear system identification is the determination of system order. The use of feedforward neural network for nonlinear system order and neural network is closely related to the promotion of the problem. OLS algorithm is a kind of learning algorithm to construct radial basis neural network, but there is a general problem to construct neural network using OLS algorithm. The ROLS algorithm combines the OLS algorithm with the regularization method to improve the generalization ability of the algorithm. In this paper, Radial Basis Network (ROLS) algorithm is applied to the order of nonlinear system. In this paper, the proposed method is simulated and the results verify the effectiveness of the proposed method.