Preprocessing method based on adaptive filtering and restricted Boltzmann machine, terminal and readable storage medium

Abstract

The invention provides a flight parameter data preprocessing algorithm based on an adaptive filter and a restricted Boltzmann machine. The flight parameter data preprocessing is studied from two aspects of noise removal and feature extraction, the recorded dynamic system flight parameter data are subjected to estimation processing by using the adaptive filter, and the noise in the data is removed;the recorded flight parameter data is mapped to the low-dimensional hidden layer from the input value of the high-dimensional visual layer by utilizing the strong unsupervised learning capability ofthe Boltzmann machine; and on the basis of a small amount of label information, clustering and grouping are carried out on the low-dimensional feature data by utilizing the fact that the probability of occurrence of the feature signals is greater than the probability of occurrence of the noise feature signals, thereby achieving the purpose of feature extraction of the flight parameter data. The method is high in anti-interference capability, low in requirement on the processed flight parameter data volume and simple in algorithm, denoising, feature extraction and dimension reduction processingon the aircraft parameter data are completed, and programming is easy to achieve.

Description

technical field [0001] The present invention relates to the technical field of signal processing, and in particular to an aircraft parameter preprocessing method based on adaptive filtering and a restricted Boltzmann machine of a military flight big data maintenance field autonomous support information support system, a terminal and a readable storage medium . Background technique [0002] From the 1990s to the present, aviation equipment technology has developed rapidly, especially in the context of military strategy adjustments and changes in the combat use of aviation equipment. Combat aircraft have recorded a large amount of flight parameter data in the process of use and maintenance. These data It is extremely valuable for ensuring flight safety and evaluating flight training. In order to make more effective use of flight parameter data, data warehouse and data mining techniques have been applied to the management and processing of flight parameter data. However, the ...

AI Technical Summary

Problems solved by technology

[0004] On the other hand, the flying parameter data contains a large number of parameters, and the relationship between these parameters is not clear. If the required parameters are not selected for the mining task, the mining workload will inevitably be huge, and the mining results cannot be guaranteed. quality

Images