弱测量问题的信息处理方法初探

Preliminary Investigation on Information Processing for the Weak Observation Problem

当前越来越多的探测系统面临由低信噪比、低数据率、低分辨率和低测量维度产生的弱测量问题。该文剖析了弱测量问题的由来及其信息集合的困难,阐述了利用探测对象本身存在的"重复性"和"可预见性"解决信息集合的思路,并提出利用贝叶斯理论处理状态估计等弱测量问题的"概率云"推演方法,最后,在信息系统设计层面探讨了解决弱测量问题导致在传感器设计、信息处理和系统控制等3方面的新变化。

Increasingly, detecting systems are facing the common problem of information processing owing to low SNR, low data rate, low resolution, and low information dimensions, and it is called the weak observation problem. This paper analyzes its origin and proposes the concept of information assembling by using the repetition and prediction properties of the object of interest. Then, the probability of the cloud inference method based on Bayesian theory is proposed to address a weak observation problem such as state estimation. Eventually several new requirements for sensor design, information processing, and system control are discussed, which are three crucial factors in information system design.