| Fault diagnosis based on knowledge is an effective method,mainly through the theoretical analysis of knowledge to realize the fault diagnosis,and it can be used to detect,extract and process the fault data of coal mine scraper conveyor.Scraper conveyor as an important link between the working and the outside world,and plays a very important position in the process of coal mining.With the development of technology,the scraper conveyor has been developed to the heavy and automation type.Due to the bad working environment,the faults are easily happened in the components of the large heavy scraper conveyor,mainly for the transmission section failure,which is the reducer,motor and hydraulic coupler failure.SGZ1250/2565 scraper conveyor is as an example to analyze the reasons for the failure of the transmission section,and its abnormal phenomenon in the early stage of fault is studied.Because the early abnormal fault signal is weak,so it can not be found in time and may cause a serious fault,even causes accidents of mine production.Based on knowledge the fault diagnosis method is put forward to analysis and diagnosis fault.In the coal mine working face operation,there are a variety of machines to work together,therefore the noises are much more serious,and the abnormal signal amplitude is relatively small in the early stage of fault,it is very difficult to collect the fault data of the scraper conveyor.Based on chaos theory,the chaotic Duffing oscillator mathematical model is established to detect the weak signal of abnormal fault,then,the feature extraction method of decomposition and reconstruction of the second generation wavelet transform is used to extract the characteristic signal in the severe noise background.There are many kinds of sensors used in the fault monitoring equipment and monitoring points of the transmission section,but the single sensor detection data analysis may produce a large error.Based on the multi-sensor data fusion theory,the reducer fault diagnosis model is established,and the radial neural network and fuzzy integral algorithm are used to improve the fault tolerance of system fault diagnosis.In the process of transmission information processing,the fault data have the characteristics of nonlinear small sample.In order to accurately analyze the fault types,the support vector machine(SVM)classification method is used to determine the selection criteria of kernel function of support vector machines.For better classification results,the rough set theory is used to reduce the redundant information,so that the fault data samples processed by SVM and rough set theory are more accurate.According to the failure mode of the scraper conveyor speed reducer transmission section,which has mutual interference,the uncertainty of the fault point and the fuzzy diversity of the fault symptom,the artificial neural network and quantum computing theory are combined to establish the quantum neural network fault classification model that is based on quantum rotation gates,and the fault analysis method of the quantum neural network with multi excitation function is proposed to improve the fault diagnosis.Based on the above analysis,the fault monitoring system of the scraper conveyor transmission section is designed,and the theoretical analysis method is used to deal with the monitoring data and the fault types,then the outcome are compared with the result of maintenance to verify the accuracy of the monitoring system.The results show that the online monitoring system is consistent with the actual fault of the scraper conveyor.It is proved that the system is reliable and strong applicability,therefore,the fault monitoring system can be used in the fault diagnosis of the transmission section of the coal mine scraper conveyor,and it can reduce the maintenance cost and improve the economic benefit.Therefore the fault monitoring system is helpful to make the maintenance decision and life prediction of the scraper conveyor. |
PDF Full Text Request