Alarm monitoring method and system, electronic equipment and computer readable storage medium

Abstract

The invention provides an alarm monitoring method and system, electronic equipment and a computer readable storage medium, and relates to the technical field of monitoring, the method comprises the following steps: obtaining current waveform data, the current waveform data being current monitoring data arranged according to a time sequence; and determining a target alarm result according to the current waveform data and the trained regression alarm model. When a regression alarm model is trained, expert knowledge and a data rule of waveform data are fused, a system can accept non-strict assertion, and the input requirement of the system is reduced; industry experts can design some expert knowledge to start the system without thoroughly knowing the whole system, so that preparation before starting is reduced, and time consumed for system development is shortened.

Description

technical field [0001] The present invention relates to the technical field of monitoring, in particular to an alarm monitoring method, system, electronic equipment and computer-readable storage medium. Background technique [0002] The alarm monitoring system based on waveform data is used for timing detection of user services. When the system is running, it automatically alarms according to the waveform data input by the user. The waveform data is monitoring data arranged in time sequence. In the alarm monitoring system, there are usually three models as follows: [0003] 1. A supervised time series model, that is, model learning is performed by inputting a piece of historical waveform data and its alarm flags. When predicting, input the current waveform data in the model and output the corresponding alarm flags; [0004] 2. Unsupervised time series model, that is, model learning is performed by inputting a piece of historical waveform data, and the current waveform data ...

AI Technical Summary

Problems solved by technology

[0006] However, the above three models all have shortcomings: 1. The supervised time series model relies on labeling, high labor costs, data analysis is prior to project development, strictly depends on business rules, and has poor versatility; 2. The unsupervised time series model relies on experts in the early stage. The ability to grasp the business, the development cycle is long, the effect is slow, and the late stage relies on the expert's ability to grasp the data, and the alarm accuracy / recall rate is not good; 3. The expert model relies on the summary of expert experience after business accumulation

[0007] In summary, the existing alarm monitoring schemes rely on manual labeling or expert experience, so the system input requirements are relatively high, resulting in more preparation work before the system starts, and the system development takes a long time

Images