Research On Early Accident Monitoring And Recognition Method Of Shale Gas Fracturing Pump Based On Infrared Thermal Imaging Technology

The infrared thermal imaging monitoring technology applied to shale gas fracturing field equipment will work properly under stricter conditions,which requires a higher processing speed,greater transmission efficiency and better visual presentation quality,due to the complicated situations and diverse monitoring requirements.Research in the thesis proposes methods to overcome the common difficulties in the application process of infrared thermal imaging monitoring technology for fracturing plunger pumps in production sites.The research is based on 3 parts,including abnormal data cleaning,monitoring images enhancement,and early accident monitoring and identification at outlet end.The feasibility and effectiveness have been proven by the case study.(1)Aiming at the difficulty that passing foreign objects in front of the lens affect monitoring results and cause false alarm in the process of monitoring,an abnormal infrared thermal imaging data recognition and cleaning method based on cross-border detection is proposed,which can identify the video start frame and end frame of foreign objects entering and exiting precisely,and transmit frame sequential information of abnormal monitoring data to the cleaning system.The computational complexity will tumble significantly by the discriminant index based on the similarity of numerical curve of sequential one-dimensional vector(up to 98.79%)to improve the recognition efficiency.After cleaning and reconstruction,the temperature changes of adjacent frames tend to be continuous and stable,eliminating the interference of abnormal objects to the monitoring process.The reconstructed data can maintain the continuity of temperature change based on infrared thermal imaging observation,which will avoid data distortion and key information loss caused by over-cleaning.(2)Aiming at the low level of local contrast and detail information visualization in infrared thermal images of complex systems,an image enhancement method based on gray value distribution optimization is proposed.Five parts or regions are selected as the research subjects,including pump head,power end,electric motor,transmission case and inlet&outlet end.Three indexes,namely average gray value temperature ratio,gray value range and principal gray level displacement,are defined as evaluation criterions.The visual effects of infrared thermal images and the visual identifiability of parameters changes before and after enhancement processing of monitoring objects in different stages are compared and analyzed.The results show that the proposed image enhancement method has obvious optimization effect on the components or regions with small temperature variation and large difference between local temperature range and overall temperature range(After treatment,the maximum increase of average gray value temperature ratio,gray value range and principle gray level displacement were 6.36 times,2.88 times and 3.33 times respectively),which will effectively improve the visualization level of detail information in infrared thermal imaging monitoring.(3)Aiming at the monitoring and identification of early leakage accidents at the outlet end of shale gas fracturing pump,a method of identifying and judging abnormal temperature change based on the distribution of RGB value of infrared thermal image is proposed.It will reduce averagely the minimum distinguishable temperature difference of images to 29.41% of that before processing,and averagely cut down the corresponding temperature division from 0.34℃ to 0.10℃ for each RGB value,which shows that proposed method can significantly optimize the color distinguishability of unit temperature difference and improve the image visualization level of surface temperature difference or change of equipment.Based on the statistical analysis of the RGB spatial coordinates and the number of pixels,the differences in the visual effects of the infrared thermal images corresponding to the surface temperature changes at the outlet end can be quantitatively evaluated.The one-dimensional change of surface temperature at outlet end from 29.45℃ to 27.00℃ is transformed to the three-dimensional change in RGB value space from(0.842,0,0)to(0.563,1,0.438)after the leakage accident,which will optimize the visual distinguishability of tiny parameter differences,and eliminate the effects from color perception or other subjective factors of monitoring staff,to conduce the observation,understanding and recognition in field monitoring.