应用非线性系统识别的RBF神经网络新型学习算法

A new learning algorithm for RBF neural networks with applications in nonlinear system identification

扩展了一个在线的优先权更新算法,即一个基于RBF神经网络的非线性不连续时间多元动态系统的识别技术,这种技术适合神经网络结构.描述了独立表示的在线算法的2个不同问题,通过建立识别问题和在适当的控制理论中揭示某些技术之间的连接,给出了一个能满足单一变量系统需要的算法.

It was introduced an on-line priority updating algorithm, nonlinear continuous-time dynamic systems of multi-recognition technology based on RBF neural network. The technique was showed suitable for neural network structure. The priority clarified the independent online algorithm with two different issues. Through the establishment of identification problems and by appropriate, control theory of a link between certain technology, it was created a system to meet the needs of a single variable algorithm.