Distributed deviation detection in sensor networks
ACM SIGMOD Record
Distributed regression: an efficient framework for modeling sensor network data
Proceedings of the 3rd international symposium on Information processing in sensor networks
| Hi-index | 0.00 |
With the development and application of modern science and technology, many new technical measurement methods have been put forward successively which are of high resolution and high collection rate about microseismic monitoring. We urgently need an effective detection method of abnormal data (mine earthquake) to collect lots of data to make real-time detection. In the past, we usually depend on experienced professional staff to solve this kind of problems. They make judgments by comparing numerical size or analysis change trend of factors. The key problem in this paper is how to find abnormal data automatically by linear autoregressive analysis (include gross error) and point out the position of abnormal data. Through this way, we can give prediction model and prediction mechanism of data stream in coal mine microseism. On the basic of this prediction model, we put out a detecting method of abnormal data, and we can detect whether data at this moment is abnormal by calculating the ratio of prediction error and average forecasting error at this moment. The results of the experiments show correctness and efficiency, and it indicates this model can make real-time detection of mine earthquake abnormal event.