基于神经网络的多维非线性模型参数估计

Parameter Estimation for Multi Dimension Non Linear Model via Neural Network

本文详细研究了多层ＢＰ神经网络的特性，并提出了基于预畸变处理的广义线性模型参数估计方法。借助于函数组合器，提出了基于神经网络的多维非线性模型参数估计算法，从而为多维非线性系统的参数辨识和分析提供一种有效途径。模拟结果表明。

In this paper, the characteristic of multi layer BP neural network has been studied in detail. The neural network method for parameter estimation of broadsense linear model is proposed via pre distorted technology. By means of function combiner, the method for parameter estimation of multi dimensional nonlinear models is also proposed via neural networks. A valid method for parameter identification and analysis of multi dimentional nonlinear systems is provided. The simulation results show that the parameter estimation method for multi dimensional nonlinear models is successful.