Fault matching and early warning method based on fault data state matrix

Abstract

The invention relates to a fault matching and early warning method based on a fault data state matrix. The fault matching and early warning method comprises the following steps: S1, data normalization processing: performing data standardization on the mean value and the standard deviation of original fault data by using a Z-score standardization method; S2, calculating a similarity distance of the fault data sample vector; S3, sequencing the fault data sample vectors according to the similarity distance, and selecting the corresponding number of sample vectors to form a fault sample state matrix according to a sample vector selection rule; and S4, storing the formed fault sample state matrix into a fault database. The fault matching and early warning method has the advantages that compared with a traditional early warning system, another way is developed, and a fault data state matrix is established, and the fault matching and early warning method is more efficient; the fault database is established, so that data accumulation is realized, and a greater possibility is provided for later data utilization; and the establishment of the fault data state matrix is realized by a computer, so that personal errors are greatly reduced.

Description

technical field [0001] The invention relates to a method for fault matching and early warning, more specifically, it relates to a method for fault matching and early warning based on a state matrix of fault data. Background technique [0002] The existing early warning systems all use the normal data when the equipment is running, and establish the state matrix of the normal data by filtering out the normal data under various working conditions when the equipment is running, and use the real-time data collected from the field to compare the similarity with the normal data. Calculate, establish real-time data trend graph and sample data trend graph, compare the deviation value of real-time data and sample data with the set alarm value, and generate an alarm if the alarm value is exceeded. The disadvantage of this early warning system is that it is too dependent on sample data and The selection of alarm value, if the sample data and alarm value are not selected properly, there...

AI Technical Summary

Problems solved by technology

[0002] The existing early warning systems all use the normal data when the equipment is running, and establish the state matrix of the normal data by filtering out the normal data under various working conditions when the equipment is running, and use the real-time data collected from the field to compare the similarity with the normal data. Calculate, establish real-time data trend graph and sample data trend graph, compare the deviation value of real-time data and sample data with the set alarm value, and generate an alarm if the alarm value is exceeded. The disadvantage of this early warning system is that it is too dependent on sample data and The selection of alarm value, if the sample data and alarm value are not selected properly, this early warning system will have problems

Images