| With the rapid development of information technology,more and more developed mining countries pay attention to mine informatization and automation technology,and begin to formulate long-term development plans.Therefore,the smart mine system characterized by thorough perception,and intelligent applications was proposed,advocating that data processing should be the core in intelligentization of mines.The online intelligent monitoring technology of coal caving volume in fully mechanized caving face proposed in this paper will be one of important technologies to realize the smart mine system.The laser scanning method is used to obtain high-precision point cloud data and construct a real-time model of coal caving coal flow.Dimensional accurate inversion of coal flow changes in fully mechanized caving face provides effective guidance for hydraulic support and scraper control,so as to achieve thorough perception in the smart mine system.The specific research content is as follows:According to the actual working conditions and needs of the fully mechanized caving face,a laser scanning system scheme that can be applied to the coal flow monitoring of the fully mechanized caving face is designed;In accordance with the explosion-proof requirements and standard design formulas,a set of devices with high explosion-proof performance is developed and equipped with a host computer software with functions such as graphics rendering,real-time communication and error feedback was built;The coal caving monitoring system was initially set up in the laboratory,and a multi-factor analysis of variance of the laser scanning effect was performed.Experiments show that,regardless of the interaction between various factors,the color,dust and scanning angle all have a significant effect on laser scanning.Aiming at the poor working conditions of fully mechanized caving face,a coal caving data interpolation algorithm is proposed,which combines the least square method and Kalman filter to realize the prediction of the missing point cloud data.Through the control experiment with the cubic spline interpolation algorithm,it is initially verified the effectiveness of the algorithm;Combined with the FPFH point cloud feature extraction method,the adaboost machine learning method is used to classify the point cloud image of the coal flow point cloud.The cross-validation control experiment based on F-score,shows that the coal flow recognition model has strong generalization ability and can be applied to the monitoring of coal caving in fully mechanized caving face;the outline of the unloaded scraper is extracted by the projection mapping method,According to the principle of the differential element method of definite integral,the coal volume calculation is performed on the entire point cloud image,and the measurement error is analyzed,and it is concluded that the coal volume volume algorithm based on the triangular differential element method has high robustness.In order to explore the reliability and measurement accuracy of the coal caving monitoring system,an industrial test was carried out at the 8222 working face of Tashan Coal Mine of Datong Coal Mining Group;The reliability test of the coal caving monitoring system was carried out,and the success-failure timing interception based on binomial distribution was carried out.The tail test shows that the coal caving monitoring system has a certain degree of reliability;The on-site measurement test of the coal caving monitoring system is carried out,and the high-precision load cell is used to verify the coal caving data of the system,and the test results are filtered and noise-reduced.It is concluded that the measurement relative error of the coal caving monitoring system is 6?13%,the speed of the scraper and the relative error are very positively correlated,and the real volume is not correlated with the relative error.This paper has 49 figures,21 tables and 88 references. |
PDF Full Text Request