Abstract

Using the previously developed numerical-analytic method for describing input and output parameters of a the multidimensional dynamical object for the given domain of admissibility (Part I), the paper develops a method for optimal estimating the values of continuous linear functionals (numerical characteristics) of measurable parameters based on incorrect data that contains not only fluctuation error, but also singular interference (Part II). The method provides maximally maximal decomposition of computation, does not require performing standard linearization scheme and choosing initial condition, and also is not related with spectral coefficient calculation in finite linear combinations (with given basis functions) describing integral curves of the differential equation, measured parameters, and singular interference. Casual and method errors are analyzed, and an illustrative context and recommendations for a practical application of the results are provided.

中文翻译：

---

描述和估计多维动态对象的输入和输出参数的数值分析方法：第二部分

摘要

使用先前开发的数值分析方法来描述多维动力学对象在给定的可允许性域中的输入和输出参数（第一部分），本文开发了一种用于最优估计连续线性函数（数值特征）值的方法。基于不正确数据的可测量参数，这些数据不仅包含波动误差，而且还包含奇异干扰（第二部分）。该方法提供了最大的计算分解，不需要执行标准的线性化方案和选择初始条件，并且与描述微分方程，测量参数的积分曲线的有限线性组合（具有给定的基函数）中的频谱系数计算无关。以及单一干扰。