一种基于传感器数据综合处理算法

提出基于知识参量的数据标准化处理方法,介绍了知识参量的构成以及故障征兆的隶属度表示方法.根据功率相等的原则提出了传感器数据中的野值滤波方法.运用神经网络和最小二乘法加权的方法对缺损数据进行填充,并分别与最小二乘法和BP网络的处理结果进行比较.仿真结果表明,运用该方法进行数据处理,可以在保证数据完整准确的情况下,有效地弥补原始数据的缺陷,提高了数据的准确性和可信度.

A Kind Of Synthetical Treatment Arithmetic Based On Sensor Data

This paper brings forward a standard method of data treatment,and introduces the structure of knowledge parameter and the press method of subjection for fault omen.Based on the principle of power equation,a method of filter of unruly value in sensor data is introduced.In this paper, a weight method is used to filling missing data.Compare with neural or least-squares,The methods of data treatment mentioned offer believable data source for data analysis later.It is proved that these methods of data treatment can not only ensure the veracity and integrality of data,but also remedy bugs in raw data,so they reduce errors in later works.