| RF功放非线性特性的在线神经网络建模 |
| |
| 引用本文： | 薛传宝,魏海坤,宋文忠. RF功放非线性特性的在线神经网络建模[J]. 东南大学学报(自然科学版), 2005, 0(Z2) |
| |
| 作者姓名： | 薛传宝 魏海坤 宋文忠 |
| |
| 作者单位： | 东南大学自动控制系 南京210096 |
| |
| 摘 要： | 为了改善功放系统的非线性特性,采用了基带数字预失真技术,其中功放系统的建模是其关键技术.由于功放系统是一个复杂的带记忆效应的非线性时变系统,为了能够实现实时校正,采用级联相关算法进行在线神经网络建模.首先选用残差相关性判别方法来确定神经网络功放模型的结构和初始参数,然后使用带遗忘因子的递推最小二乘法对神经网络模型参数进行在线自适应调整.对实测数据的验证表明,建立的神经网络模型完全能达到给定的性能指标要求.
|
| 关 键 词： | 神经网络 功率放大器 非线性建模 级连相关算法 递推最小二乘 遗忘因子 |
On-line neural network method for modeling RF power amplifiers nonlinear characteristic |
| |
| Xue Chuanbao Wei Haikun Song Wenzhong. On-line neural network method for modeling RF power amplifiers nonlinear characteristic[J]. Journal of Southeast University(Natural Science Edition), 2005, 0(Z2) |
| |
| Authors: | Xue Chuanbao Wei Haikun Song Wenzhong |
| |
| Abstract: | To improve power amplifiers'(PAs) nonlinear characteristic,digital baseband predistortion is used.Modeling power amplifiers is one of the key technology.The PAs system is complex,usually nonlinear,time-varying,and has memory effect.To realize real-time predistorter that performs well,cascade-correlation algorithm is used for online modeling the nonlinear characteristic of PAs.First,residual correlation method is used to determine the structure and initial parameters of the neural network model for power amplifiers system,then the model parameters are online adjusted with forgetting factor recursive least square algorithms(FFLS).Application shows that the neural network model reaches the performance index satisfactorily. |
| |
| Keywords: | neural network power amplifiers nonlinear modeling cascade-correlation recursive least squares forgetting factor |
| 本文献已被 CNKI 等数据库收录！ |
|
