Research On Shearer Cutting Pattern Recognition Method Based On Infrared Thermal Imaging

As one of the important equipments in fully mechanized coal mining face,its intelligent level is the key factor that restricts the development of “unmannedor” or “less humanization” in fully mechanized mining face,and the cutting pattern recognition is the key to realize the intelligent control of shearer.At present,shearer can't adjust its working state adaptively when the property of coal and rock changes.Frequent manual intervention is still necessary.Therefore,it's essential to study the technology of cutting pattern recognition and improve the intelligence level of shearer.In this paper,the infrared thermal imaging signal of shearer cutting the coal wall is taken as the research subject.The adaptive denosing method for infrared thermal image and the position tracking method of shearer cutting unit are studied.And the cutting pattern recognition of shearer is realized based on neural network.The main work and research results of this paper could be expounded as follows:(1)Based on the study of temperature measure principle,image-forming principle of infrared thermal imaging technology and the basic structure of shearer,the cutting pattern of shearer is analyzed.The influence factors of coal wall temperature change in shearer cutting process and the temperature distribution characteristics of the coal wall's infrared thermal image are studied.(2)The characteristics of infrared thermal image of shearer in fully mechanized mining face are analyzed.In view of the noise problem of infrared thermal image,taking advantage of the denosing ability of dual-tree complex wavelet transform and bilateral filter combined with the fast optimization ability of fruit fly algorithm,an image denosing algorithm based on duel-tree complex wavelet optimized by fruit fly algorithm and bilateral filter is studied.(3)The shape feature of shearer cutting unit in infrared thermal imaging video is analyzed,and the location method of shearer cutting unit based on morphology is studied.The tracking algorithm of shearer cutting unit based on morphology and spacetime context is proposed.The positon tracking of cutting unit in the infrared thermal imaging video is realized,which lays the foundation for obtaining the coal wall temperature befor and after cutting.(4)The temperature change rule of coal wall before and after cutting is studied.The highest temperature difference value,the minimum temperature difference value and the average temperature difference value of coal wall before and after cutting are selected as recognition characteristics.The cutting pattern recognition model of shearer based on BP neural network is set up.The parameters of the neural network are optimized.It improved the shearer cutting pattern recognition rate and provides a basis for realizing intelligent control of shearer.The experiment was tested at Zhangjiakou Coal Mining Machinery Co.Ltd.national energy mining equipment research and development center.Experimental results show that the image denoising algorithm has good denoising effect on infrared thermal imaging,the tracking algorithm can effectively track the shearer cutting unit and cutting pattern recognition method based on coal wall temperature can accurately identify shearer cutting pattern,demonstrate the correctness and effectiveness of the proposed method.