An Eigenfunction Approach to Parameter Estimation for 1D Diffusion Problems

An Eigenfunction Approach to Parameter Estimation for 1D Diffusion Problems
复制标题

一维扩散问题参数估计的特征函数方法

DOI:
--
复制
发表时间:
2019
期刊:
European Control Conference
影响因子:
--
通讯作者:
R. Rabenstein
R. Rabenstein
中科院分区:
--
文献类型:
--
作者:
Maximilian Schäfer;Alexander Ruderer;R. Rabenstein

文献摘要

被引文献

相似文献

分布参数系统的特性可以用本征函数的展开式来表示。它允许根据系统的参数计算输出信号。这一贡献认为逆问题:估计系统参数的噪声测量的输出。为此,本征函数展开用于建立分布参数系统的状态空间描述。相应的状态空间矩阵定义了一个扩展卡尔曼滤波器来执行参数值的估计和跟踪。这种方法在这里显示的扩散问题在一个空间维度。的扩散参数的值估计从不同条件下的模拟颗粒浓度。这类问题出现在例如新兴的分子通信领域中。
The behavior of a distributed parameter system can be represented by an expansion into eigenfunctions. It allows to calculate the output signal in dependence on the parameters of the system. This contribution considers the inverse problem: Estimate the system parameters from noisy measurements of the output. To this end, an eigenfunction expansion serves to establish a state-space description of the distributed parameter system. The corresponding state-space matrices define an extended Kalman filter to perform estimation and tracking of parameter values. This approach is shown here for a diffusion problem in one spatial dimension. The value of the diffusion parameter is estimated from a simulated particle concentration under varying conditions. Problems of this kind arises for example in the emerging field of molecular communications.