基于粗糙集神经网络的非线性系统动态参数预测

Dynamic Parameters Prediction of Nonlinear System Based on Rough Sets and Artificial Neural Network

为建立非线性系统辨识和预测模型,利用粗糙集和人工神经网络方法进行动态参数预测。考虑非线性系统中影响因子之间的高度非线性和不确定性,结合粗糙集和人工神经网络的优点,提出了一种动态参数预测的新方法。该方法充分考虑了学习样本的数据特性,简化了决策规则从而降低了网络拓扑结构规模,计算速度快,容错能力强,误差小,精度高。计算结果表明,该方法用于非线性系统动态参数分析是有效可行的。

The aim of research is to set up prediction model of nonlinear system. Take advantage of rough sets and artificial neural network. Consider the nonlinearity and uncertainty between the impact factors of nonlinear system. Set up the topological structure of RS-ANN. Present a new method on identification and prediction of nonlinear system dynamic parameters. Rough sets can judge the characters of sample data of ANN so get the decision rules easily. This method simplifies the structure of network and reduces the calculation time so get high speed in processing the data. It has high ability of fault-tolerant and has small errors. Results of calculation prove its efficiency and availability in the dynamic analysis of nonlinear system.