论文部分内容阅读
提出一种基于状态空间投影方程的动态多尺度系统建模方法 ,给出了基于 Haar小波的模型的具体形式。该模型满足标准卡尔曼滤波条件 ,执行卡尔曼滤波 ,可获得各个尺度上目标状态线性最小方差意义下的最优融合估计值 ,仿真结果令人满意 ,并为动态多尺度系统建立了一种新的估计理论框架。
A dynamic multiscale system modeling method based on the state space projection equation is proposed. The specific form of the model based on Haar wavelet is given. The model satisfies the standard Kalman filter conditions and performs Kalman filtering to obtain the optimal fusion estimate in the sense of the minimum linear variance of the target state on each scale. The simulation results are satisfactory and a new type of dynamic multiscale system is established The theoretical framework of estimation.