具有Hammerstein模型特征的非线性远程预报辨识

Identification of Nonlinear Remote Prediction Based on Hammerstein Model

为提高内河船舶远程通信预警功能的精度和可靠性,提出了具有非线性预测控制模型特征的远程复杂目标预报辨识算法。通过将远程复杂目标的辨识分解为一个静态非线性环节和动态线性环节的串联,利用稳态信息获取稳态模型的一致性估计,并通过动态模型获得非线性静态环节的增益,再利用奇异值分解法和动态信息辨识获取非线性系统未知参数的估计。仿真结果验证了该方法的有效性。

A predictive identification long-range target of nonlinear predictive telecommunication for the inland-river ships. algorithm is proposed, which, based on the Hammerstein model for control, improves the precision and reliability of remote The identification of complicated remote target is decomposed to a string of static nonlinear block and a dynamical linear block, thus to acquire the consistency estimation of static model by steady-state input, while to achieve the gain of static nonlinear block by dynamical input, so the estimation of unknown parameters for the nonlinear system could be obtained. The efficiency of identification algorithm is verified by simulation results.