Complex Theory and Batch Processing in Mechanical Systemic Data Extraction

Abstract

This paper designs a new batching program to extract the original data, which helps to traverse the entire sample space quickly and provides a new approach for data extraction based on the motion stroke diagram. The designed program can read thousands of files out of many folders instantly and automatically. Meanwhile, thousands of time nodes are calculated based on the proportional coefficients. Finally, experimental data of many folders are separated into sample space easily and rapidly. The original data, the extracted stroke data, and the name and address of every folder and file are output in the result. The program designed in this paper at the maximum processing speed needs only 0.015 seconds to read, compute, and extract the correlative data information from one file (s/f), and the average time threshold is 0.0866 seconds. The Linear Theory (LT), optimizing Sparrow Search Algorithm (OSSA), and the automatic batch read file method can be employed to obtain the optimal result of data extraction. Through 744 rounds of nine experiments, the Average Processing Speed (APS) is less than 0.110 seconds per second data segment (s/ess). The APS is increased by 79.73%. The accuracy of fault classification and forecast of the eigenvalue extracted from the Automatic Batch Reading File (ABRF) and the Ensemble Empirical Mode Decomposition (EEMD) method is improved by 9% and 13% by Self-Organized Mapping (SOM). It is validated that our proposed data extraction method is faster and more progressive than the existing ones.

Keywords

• Automatic batch reading file
• batch processing
• data extraction
• motion stroke
• processing speed