Research On Slippage Detection Method Of Elevator Traction Wheel Based On Image Processing

With the continuous development of social economy and the increasing number of high-rise buildings,elevators have become an indispensable equipment in modern buildings,among which traction elevators are most commonly used.The traction elevator mainly relies on the friction between the traction sheaves and the wire ropes to realize its up and down operation.Adequate towing capability is the basis for ensuring safe operation of elevators,and the slippage then is an significant indicator go towards evaluating the towing ability.However,neither the present national standard nor the safety technical standard has put forward a definite quantitative method for the condition of wear and tear of the lead wheel,between the tow wheel and the wire rope.Therefore,it is of great significance to study the test method of the elevator’s towing wheel slip which may have a scientific judgement on the failure of the towing wheel.Meanwhile,it also has become a hot issue in the field of elevator safety.On another point,due to the advantages of non-contact,high precision,high flexibility and applicability,the image processing technology is widely used in industry,agriculture,medicine and the military.On the basis of the technical principle of image processing,the mechanism and testing method of the elevator towing wheel and steel wire rope sliding was studied and a method based on that was proposed as well.The major studies completed were as follows:1)Overall scheme design of elevator traction wheel slip detection.n the analysis of the characteristics and the detection precision of the elevator traction wheel,the general scheme design included hardware,software,algorithm and test.The hardware part included hardware selection and platform construction;software part included image acquisition software and image processing and development software,algorithm part included image transformation and core method to detect the slip of elevator traction wheel,and test part included simulation and live machine testing.2)Research on geometric transformation algorithm of sliding image of traction wheel.Because the traction wheel is circular,the geometric transformation algorithm of the traction wheel image is studied in order to transform the elevator traction wheel slip image with different object distance on each part of the traction wheel into the same object distance.Firstly,this paper analyzed the rules of geometric transformation such as translation,mirror image,transposition,rotation and dissecting transform of image,and presented an improved geometric transformation rule applied to the original image of the mark based on the principle that the pixel point was equal to the arc length of the center position and the line segment length of the target image.Then the three common gray scale filling algorithms were compared,and the linear interpolation algorithm was selected for grayscale based on the characteristics of short running time and good fill effect.Finally,the feasibility of the geometric transformation algorithm of this image was verified by simulation test.3)Research on slip image detection algorithm of elevator traction wheel.In order to conduct the quantitative inspection of elevator towing wheel slip,we first carried out the calibration of pixel equivalence and the calibration of camera,then analyzed the principle of gray scale calculation of image and the gray scale image was obtained by weighted average method.Secondly,the effect of different denoise algorithms was analyzed by observing and comparing the characteristics of the image of the markers,and the medium value filtering was chosen to denoise.The region growth algorithm would be used as the segmentation algorithm in this paper,the image of the object was obtained after dividing the wire rope by opening operation,and the sliding quantity of the elevator’s towing wheel was determined by mass core method.4)Elevator traction wheel slip test and error analysis.In order to verify the performance of the elevator traction wheel slip detection system,many times in Huzhou special equipment research institute elevator training center field test.Test result of single marked solid machine: The absolute error was 0.74 mm,the relative error was 2.2%,the spreading uncertainty of the slip detection was 33.8 ± 0.69 mmand the confidence probability P=0.95 and the degree of freedom v=8.In order to solve the problem of excessive slip in a single marker test,which resulted in a "coil" of the marker,a multi-marker detection test was performed and a multi-marker solid-machine test was proceeded in a small area of Huzhou.The test result was: the absolute error was 0.52 mm and the relative error was 3.91 percent.To sum up,this paper studied the method of testing the elevator’s towing wheel by using improved image geometry transform algorithm to convert the object-distance image to the same target image.The results of the test showed that the system could meet the precision requirements.