Gas monitoring data anomaly identification based on spatio-temporal correlativity analysis

Based on spatio-temporal correlativity analysis method, the automatic identification techniques for data anomaly monitoring of coal mining working face gas are presented. The asynchronous correlative characteristics of gas migration in working face airflow direction are qualitatively analyzed. The calculation method of asynchronous correlation delay step and the prediction and inversion formulas of gas concentration changing with time and space after gas emission in the air return roadway are provided. By calculating one hundred and fifty groups of gas sensors data series from a coal mine which have the theoretical correlativity, the correlative coefficient values range of eight kinds of data anomaly is obtained. Then the gas moni- toring data anomaly identification algorithm based on spatio-temporal correlativity analysis is accordingly presented. In order to improve the efficiency of analysis, the gas sensors code rules which can express the spatial topological relations are sug- gested. The experiments indicate that methods presented in this article can effectively compensate the defects of methods based on a single gas sensor monitoring data.

Integrated Generative Adversarial Network and XGBoost for Anomaly Processing of Massive Data Flow in Dispatch Automation Systems

Existing power anomaly detection is mainly based on a pattern matching algorithm.However,this method requires a lot of manual work,is time-consuming,and cannot detect unknown anomalies.Moreover,a large amount of labeled anomaly data is required in machine learning-based anomaly detection.Therefore,this paper proposes the application of a generative adversarial network(GAN)to massive data stream anomaly identification,diagnosis,and prediction in power dispatching automation systems.Firstly,to address the problem of the small amount of anomaly data,a GAN is used to obtain reliable labeled datasets for fault diagnosis model training based on a few labeled data points.Then,a two-step detection process is designed for the characteristics of grid anomalies,where the generated samples are first input to the XGBoost recognition system to identify the large class of anomalies in the first step.Thereafter,the data processed in the first step are input to the joint model of Convolutional Neural Networks(CNN)and Long short-term memory(LSTM)for fine-grained analysis to detect the small class of anomalies in the second step.Extensive experiments show that our work can reduce a lot of manual work and outperform the state-of-art anomalies classification algorithms for power dispatching data network.

Comparative study on isolation forest, extended isolation forest and generalized isolation forest in detection of multivariate geochemical anomalies

It is not easy to construct a model to describe the geochemical background in geochemical anomaly detection due to the complexity of the geological setting.Isolation forest and its improved algorithms can detect geochemical anomalies without modeling the complex geochemical background.These methods can effec-tively extract multivariate anomalies from large volume of high-dimensional geochemical data with unknown population distribution.To test the performance of these algorithms in the detection of mineralization-related geochemical anomalies,the isolation forest,extended isolation forest and generalized isolation forest models were established to detect multivariate anomalies from the stream sediment survey data collected in the Wu-laga area in Heilongjiang Province.The geochemical anomalies detected by the generalized isolation forest model account for 40%of the study area,and contain 100%of the known gold deposits.The geochemical anomalies detected by the isolation forest model account for 20%of the study area,and contain 71%of the known gold deposits.The geochemical anomalies detected by the extended isolation forest algorithm account for 34%of the study area,and contain 100%of the known gold deposits.Therefore,the isolation forest mo-del,extended isolation fo-rest model and generalized isolation forest model are comparable in geochemical anomaly detection.