A filling method for missing value of coal mine gas concentration based on time-space fusion

Abstract

The invention belongs to the field of data processing, and specifically relates to a method for filling missing values ​​of coal mine gas concentration based on time-space fusion. The method includes: identifying and processing noise data; filling missing values ​​of data based on time dimension; The missing value of the data is filled; the filling results of the time and space dimensions are fused to obtain the final result; the present invention uses the offline data as a training set for model construction, and applies the generated model to the data in the production environment, which can synthesize Use various coal mine gas concentration monitoring data to improve the generalization ability of the model and reduce the filling error.

Description

technical field [0001] The invention belongs to the field of data processing, and in particular relates to a method for filling missing values ​​of coal mine gas concentration based on time-space fusion. Background technique [0002] With the continuous update and development of science and technology, informatization is becoming more and more important in various fields, from the initial computer, camera to home computer, smart phone, and then to big data and artificial intelligence, through continuous upgrading of the way of collecting and utilizing data, Connect people and people, people and the world to form a dense network, and data mining has been widely used in various fields. In practice, "mis-recording" and "missing record" may occur during the data entry process, resulting in noisy or incomplete data. Incomplete data will have multiple impacts on data mining: the reduction of complete data and the corresponding reduction of available information will lead to large...

AI Technical Summary

Problems solved by technology

During the coal mining process, sensors of different specifications are installed to collect monitoring data in different regions, different mining areas, and at different times. By mining these data, its potential value can be discovered, which plays an important role in safety monitoring and disaster early warning in mining areas; However, in practice, the collected data has problems such as noisy data and missing data and cannot be used directly. Analysis based on unprocessed data may cause large deviations and obtain wrong analysis results of coal mine safety situation, which seriously endangers people. Life and Property Safety

Most of the existing data preprocessing methods use related technologies in the fields of statistics, machine learning, and deep learning, and most missing value filling methods are only suitable for situations where the missing rate is low and the data changes smoothly. Once the missing rate is high, the filling effect will be reduced. will drop sharply, and it is difficult to use a model to fit all missing situations

Images