Research On Straightness Control Of Fully Mechanized Coal Face Based On Vision Measurement

Coal is still the main energy in China for a long time.At present,China's coal mine safety production situation continues to improve,but the mining technology conditions are becoming more and more complex,many safety factors,accidents still occur,seriously threatening the lives of coal miners.It is an important means to reduce casualties and ensure safety to improve the scientific and technological level of coal production through intelligent mining.In the face of the complex geological environment of coal seams hundreds of meters or even over kilometers deep underground,the coal mining equipment needs to constantly position and adjust the direction of advance in order to adapt to the fluctuation of coal seam level and height.The production process of coal face requires"three straight,two flat and two unblocked".In the process of continuous advancement of coal face,the shearer,hydraulic support and scraper are pushed forward together to accurately measure the straightness of the fully mechanized coal face and effectively adjust and control it.It is an important research topic for the realization of automation,intelligence and unmanned in the fully mechanized coal face.At present,the existing displacement Jack travel sensor can detect the displacement value of the support's displacement cylinder,but it can't know the absolute position of the working face where the hydraulic support itself is.At the same time,there is no effective means to detect the straightness of the working face,and the control technology is still blank.Therefore,it is particularly important to research and develop the straightness sensing and control equipment with advanced level,high reliability,modern high-yield and efficient mine construction needs and independent intellectual property rights.To achieve automation?unmanned mining,supported by national high-tech research and development plan(863 plan)funded project "coal mining technology and equipment of intelligent mining" sub-topic "fully mechanized mining intelligent control technology and equipment"(2013AA06A410)?This paper in view of the environmental impact of fully mechanized coal mining face,including insufficient lighting,coal dust and complex equipment,the paper studies the technology of finding and directing complex working face in order to provide theoretical support for the straightness control of fully mechanized mining face.The main research work is to complete the system include:1)According to the visual measurement technology,the uniform scattering LED light source with continuously adjustable luminous intensity can be designed under the dark environment or the backlighting environment of the fully mechanized mining face,which can effectively reduce the external factors such as coal dust Interference,and through a variety of color changes in the working status of the device feedback,according to different environments highlight the difference between the measured target and the background,effectively improve the quality of the measured images,and thus simplify the image processing algorithms.2)The trapezoid window matching technology is designed by using the image features of camera matching with scraper conveyor and cable trough.Starting from the image segmentation theory of gray threshold,the straightness of complex fully mechanized working face is simplified as the guard board detection of scraper conveyor cable trough edge,so as to extract the measured edge accurately and quickly,simplify the processing algorithm and control complexity.3)A new method of stroke measurement is proposed.The accuracy of stroke measurement is improved by using the opposite type ultrasonic sensor.At the same time,the "terminal control" mode of wireless communication module control integrated measuring device is developed,which makes the device easier to be arranged and the communication mode diversified.4)The straightness algorithm of monocular vision and straightness algorithm of multi vision are put forward.The collected image is matched with feature points to eliminate mismatching feature points and increase corner point detection to match the image.Finally,the image splicing process is completed under the function of perspective matrix.5)To verify the accuracy of the algorithm through field experiments.