Intelligent Monitoring of Complex Technical Systems: Application of the Forecast Aggregation Approach

The paper presents a description of an approach to intelligent analysis of the state of complex technical systems. The approach includes extracting time series from technical system logs and determining abnormal values based on forecasting results. An aggregation method for obtaining a forecast using a neural network is used. A conclusion is made about the state of the object's characteristics by integrating the results of anomaly detection and knowledge about the subject area from the OVL ontology as a result of the analysis. The proposed approach is implemented in an information system. Experiments were conducted on hardware storage system data. The approach has proven its effectiveness.