机构地区: 黑龙江大学自动化系
出 处: 《科学技术与工程》 2007年第6期948-954,共7页
摘 要: 对含未知模型参数和噪声统计的多传感器单输入单输出系统,用现代时间序列分析方法,基于自回归滑动平均(ARMA)新息模型的在线辨识,可得到未知模型参数和噪声统计估值器,进而在按状态分量标量加权线性最小方差最优信息融合准则下,提出了自校正分量解耦信息融合Wiener状态预报器。它实现了自校正分量解耦局部Wiener状态预报器和自校正分量解耦融合预报器。证明了它的收敛性和渐近最优性。一个目标跟踪系统的仿真例子说明了其有效性。 For muhisensor single input-single output systems with unknown model parameter and noise statistics, using the modem time series analysis method, based on the on-line identification of the autoregressive average moving (ARMA) innovation model, the estimators of unknown model parameters and noise statistics can be ob- tained. Further, under the linear minimum variance optimal information fusion criterion weighted by scales for state components, a self- tuning component decoupled fused Wiener state predictor is presented. It realizes the self- tuning decoupled local Wiener predictors for components and a self-tuning decoupled fusion prediction for components. Its convergence and asymptotic optimality are proved strictly. A simulation example for a target tracking system shows its effectiveness.