Dynamic threshold anomaly detection method and system, storage medium and intelligent equipment

Abstract

The invention discloses a dynamic threshold anomaly detection method and system, a storage medium and intelligent equipment, and the method comprises the steps: obtaining historical data and current data of a monitored business index, and carrying out the data cleaning operation, so as to obtain cleaning data; determining a sequence type of the current service data according to the cleaning data;inputting cleaning data into a corresponding prediction model according to the sequence type of the current service data to obtain a threshold reference value of a to-be-monitored time point; and inputting the current real data into an abnormality detection module, and when the real value of the monitoring time point exceeds a dynamic threshold range, marking the real value to be abnormal, replacing the real value with a threshold reference value, and adding the real value into a historical data queue so as to correct the model. According to the method, the fluctuation early warning accuracy and the threshold adaptability in the service monitoring process can be improved, the manual maintenance cost is saved, and the actual application requirements are met.

Description

technical field [0001] The present invention relates to the technical field of time series anomaly detection, in particular to a dynamic threshold anomaly detection method, system, storage medium and intelligent equipment. Background technique [0002] With the continuous development of telecommunication services and big data technology, the service categories and business indicators of operators continue to increase, and the scale of business data is also growing rapidly. Therefore, in the face of abnormal data fluctuations, maintaining system stability is particularly important, and related anomaly detection methods are also continuously updated. [0003] At present, the commonly used business application monitoring system usually sets relevant application thresholds to realize active early warning before data center servers and business application systems fail. Among them, for the setting of relevant application thresholds, the current mainstream method is to set static...

AI Technical Summary

Problems solved by technology

However, the threshold usually relies on manually defining the empirical threshold, and compares the data at the current moment with the previous moment. If the upper and lower limits of the threshold are exceeded, an alarm will be generated. If the threshold is too large, false positives will easily occur, and if the threshold is too small, more false positives will be generated. Therefore, the adaptability of the threshold is poor, the monitoring field of view is limited, and the cost of manual maintenance is high

Images