Method for monitoring blockage of feeding cavity of crusher

Abstract

The invention discloses a monitoring method for material blockage of a feeding cavity of a crusher, which comprises the following steps: acquiring an image of a feeding port, an image of a discharging port and an image of a discharging port of a material cavity in real time, and respectively extracting the outline of a material in the image of the feeding port, the outline of a material in the image of the discharging port and the outline of a material in the image of the discharging port of the material cavity; the unit flow Sin of the materials at the feeding port is calculated according to the outline of the materials in the image of the feeding port, the unit flow Sout of the materials at the discharging port is calculated according to the outline of the materials in the image of the discharging port, delta S = alpha * Sin-Sout, and the material level d of the materials in the material cavity is calculated according to the outline of the materials in the image of the discharging port of the material cavity; and calculating the sum delta Ssum of delta S in a preset time, and if delta Ssum is greater than a first preset value and d is greater than a second preset value, giving an alarm. Therefore, the blockage condition of the feeding cavity of the crusher can be timely found and pre-warned.

Description

Technical Field

[0001] The present invention belongs to the technical field of monitoring, and particularly relates to a method for monitoring material blockage in a crusher feeding chamber. Background Art

[0002] In actual mine exploitation and production, large stones after exploitation need to be transported to a crusher by a feeder for crushing operation.

[0003] When the crusher performs material crushing, large pieces of raw ore are transported to the crusher feeding port through a conveyor belt and poured into the crushing chamber for material crushing. The crushed ore with small particle size is transported to other equipment such as semi-autogenous mills by other conveyor belts for subsequent production. During the entire ore crushing process, factors such as too fast feeding, too large ore particle size, and too much accumulated material in the chamber may all cause blockage of the crushing chamber, thereby damaging the equipment, causing economic losses, and endangering production safety. Therefore, it is extremely necessary to timely detect and warn of the blockage of the crusher feeding chamber. Summary of the Invention

[0004] The present invention aims to at least solve one of the technical problems in the related art to some extent. For this purpose, an object of the present invention is to propose a method for monitoring material blockage in a crusher feeding chamber, aiming to be able to timely detect and warn of the blockage situation of the crusher feeding chamber.

[0005] The present invention proposes a method for monitoring material blockage in a crusher feeding chamber. According to an embodiment of the present invention, the crusher includes a feeding port, a discharging port, and a material chamber connecting the feeding port and the discharging port, and is characterized in that the method includes:

[0006] S1. Real-time obtain images of the feeding port, the discharging port, and the material chamber discharging port, and respectively extract the contours of the materials in the images of the feeding port, the contours of the materials in the images of the discharging port, and the contours of the materials in the images of the material chamber;

[0007] S2. Calculate the unit flow rate S of the material at the feeding port according to the contour of the material in the image of the feeding port in , calculate the unit flow rate S of the material at the discharging port according to the contour of the material in the image of the discharging port out , ΔS = α·S in - S out , is the flow rate of the material at the feeding port per unit time, is the flow rate of the material at the discharging port per unit time, and calculate the material level d of the material in the material chamber according to the contour of the material in the image of the material chamber;

[0008] S3. Calculate the sum ΔS of ΔS within a preset time sum , if ΔS sumIf a is greater than the first preset value and d is greater than the second preset value, an alarm is issued.

[0009] According to the monitoring method of the above embodiments of the present invention, by acquiring the images of the feeding port, the discharging port, and the material chamber in real time, the contours of the material in the image of the feeding port, the contours of the material in the image of the discharging port, and the contours of the material in the image of the material chamber can be extracted. Then, according to the contour of the material in the image of the feeding port, the unit flow rate S of the material at the feeding port is calculated. in According to the contour of the material in the image of the discharging port, the unit flow rate S of the material at the discharging port is calculated. out ΔS = α·S in -S out , is the flow rate of the material at the feeding port per unit time, is the flow rate of the material at the discharging port per unit time. By using the density ratio parameter of the incoming and outgoing materials to correct the volume change before and after the material pile is broken, the accuracy of estimating the incoming and outgoing material amounts of the crusher is improved. At the same time, the interference of the volume change caused by material crushing on the statistics of the difference between the incoming and outgoing material amounts is effectively reduced. In addition, the method provided in this application combines the warning of the material amount difference and the analysis of the material level warning, which can effectively reduce false detection and improve the accuracy of blockage identification. Thus, the blockage situation of the feeding chamber of the crusher can be detected and warned in time.

[0010] In addition, the monitoring method according to the above embodiments of the present invention may have the following additional technical features:

[0011] In some embodiments of the present invention, the Gaussian mixture model and the image difference method are used to extract the contour of the material in the image of the feeding port. Thus, based on the background modeling method for detecting and segmenting the contour of the material in the image of the feeding port, the robustness of the material contour detection algorithm can be improved, and further the accuracy of blockage identification can be improved.

[0012] In some embodiments of the present invention, the Gaussian mixture model and the image difference method are used to extract the contour of the material in the image of the discharging port. Thus, based on the background modeling method for detecting and segmenting the contour of the material in the image of the discharging port, the robustness of the material contour detection algorithm can be improved, and further the accuracy of blockage identification can be improved.

[0013] In some embodiments of the present invention, the Gaussian mixture model and the image difference method are used to extract the contour of the material in the image of the material chamber. Thus, based on the background modeling method for detecting and segmenting the contour of the material in the image of the material chamber's discharging port, the robustness of the material contour detection algorithm can be improved, and further the accuracy of blockage identification can be improved.

[0014] In some embodiments of the present invention, the crusher feeds materials into the feeding port through a first conveyor belt and discharges materials from the discharging port through a second conveyor belt. The image of the feeding port shows the materials on the first conveyor belt and in the feeding port, and the image of the discharging port shows the materials on the second conveyor belt and in the feeding port.

[0015] The Gaussian mixture model is used to obtain the image of the first conveyor belt. The image difference method is adopted, and based on the image of the first conveyor belt and the image of the feeding port, the contour of the materials in the image of the feeding port is extracted. Thus, based on the background modeling method for detecting and segmenting the contour of the materials in the image of the feeding port, the robustness of the material contour detection algorithm can be improved, and further the accuracy of blockage identification can be enhanced.

[0016] In some embodiments of the present invention, the Gaussian mixture model is used to obtain the image of the second conveyor belt. The image difference method is adopted, and based on the image of the second conveyor belt and the image of the discharging port, the contour of the materials in the image of the discharging port is extracted. Thus, based on the background modeling method for detecting and segmenting the contour of the materials in the image of the discharging port, the robustness of the material detection algorithm can be improved, and further the accuracy of blockage identification can be enhanced.

[0017] In some embodiments of the present invention, the Gaussian mixture model is used to obtain the image of the part of the material cavity not covered by materials. The image difference method is adopted, and based on the image of the part of the material cavity not covered by materials and the image of the material cavity, the contour of the materials in the image of the material cavity is extracted. Thus, based on the background modeling method for detecting and segmenting the contour of the materials in the image of the material cavity, the robustness of the material contour detection algorithm can be improved, and further the accuracy of blockage identification can be enhanced.

[0018] In some embodiments of the present invention, a coordinate system is established with a vertex in the image of the feeding port as the origin, the width direction of the first conveyor belt as the abscissa X-axis, and the direction perpendicular to the X-axis as the ordinate Y-axis. The contour of the materials in the image of the feeding port is segmented along the Y-axis. The absolute value of the coordinate of the materials in the image of the feeding port closest to the origin on the Y-axis is the minimum ordinate y min , and the absolute value of the coordinate of the materials in the image of the feeding port farthest from the origin on the Y-axis is the maximum ordinate y max , and the average proportion P min of the material width within the ordinate interval [y max is calculated: in :

[0019]

[0020] p i = h in，i / H in，i , where h in，i is the width of the materials at the i-th position in the contour of the materials in the image of the feeding port, and Hin，i In the image of the feeding port, it is the width of the first conveyor belt corresponding to the i-th position in the outline of the material;

[0021] w′ in = P in ·w in ,w in is the width of the first conveyor belt in the image of the feeding port. Calculate S based on the width w′ of the outline of the material in the image of the feeding port in Calculate S in . Thus, based on the background modeling method, the outline of the material in the image of the feeding port is detected and segmented, and it is proposed that the material cross-section model can accurately estimate the material flow density of the feeding port.

[0022] In some embodiments of the present invention, taking a vertex in the image of the discharging port as the origin, the width direction of the second conveyor belt as the abscissa X-axis, and the direction perpendicular to the X-axis as the ordinate Y-axis to establish a coordinate system, the outline of the material in the image of the discharging port is segmented along the Y-axis. The absolute value of the coordinate of the material in the image of the discharging port close to the origin on the Y-axis is the minimum ordinate y′ min , and the absolute value of the coordinate of the material in the image of the discharging port far from the origin on the Y-axis is the maximum ordinate y′ max , calculate the average proportion P of the material width within the ordinate interval [y′ min , y′ max : out :

[0023]

[0024] p′ i = h out，i / H out，i , h out，i is the width of the material at the i-th position in the outline of the material in the image of the discharging port, and H out，i is the width of the conveyor belt corresponding to the i-th position in the outline of the material in the image of the discharging port;

[0025] w′ out = P out ·w out , w out is the width of the second conveyor belt in the image of the discharging port. Calculate S based on the width w′ of the outline of the material in the image of the discharging port out Calculate S out . Thus, based on the background modeling method, the outline of the material in the image of the discharging port is detected and segmented, and it is proposed that the material cross-section model can accurately estimate the material flow density of the feeding port.

[0026] In some embodiments of the present invention, between steps S1 and S2, it further includes:

[0027] Calculate the value of m in / M in for the nth frame of the inlet image, where m in is the area of the material image in the nth frame of the inlet image, and M in is the area of the nth frame of the inlet image;

[0028] If the value of m in / M in for the nth frame of the inlet image is greater than the area threshold T a , and the difference between the value of m in / M in for the nth frame of the inlet image and the value of m in / M in for the (n - 1)th frame of the inlet image is greater than the mutation threshold T s , then repeat step S1. Thus, the robustness of the material contour detection algorithm can be effectively improved through the background update strategy.

[0029] In some embodiments of the present invention, between step S1 and S2, it further includes:

[0030] Calculate the value of m out / M out for the nth frame of the outlet image, where m out is the area of the material image in the nth frame of the outlet image, and M out is the area of the nth frame of the outlet image;

[0031] If the value of m out / M out for the nth frame of the outlet image is greater than the area threshold T′ a , and the difference between the value of m out / M out for the nth frame of the outlet image and the value of m out / M out for the (n - 1)th frame of the outlet image is greater than the mutation threshold T′ s , then repeat step Sl. Thus, the robustness of the material contour detection algorithm can be effectively improved through the background update strategy.

[0032] In some embodiments of the present invention, between step S1 and S2, it further includes:

[0033] Calculate the value of m f / M f for the nth frame of the cavity image, where m f is the area of the material image in the nth frame of the cavity image, and M f is the area of the nth frame of the cavity image;

[0034] If the value of m f / Mf The value is greater than the area threshold T″ a , and the m of the nth frame of the cavity image f / M f value and the m of the (n - 1)th frame of the cavity image f / M f The difference in values is greater than the mutation threshold T″ s , then repeat step S1. Thus, the robustness of the material contour detection algorithm can be effectively improved through the background update strategy.

[0035] In some embodiments of the present invention, step S3 further includes:

[0036] If ΔS sum is less than or equal to the first preset value and d is greater than the second preset value, a warning is issued;

[0037] If ΔS sum is greater than the first preset value and d is less than or equal to the second preset value, a warning is issued.

[0038] Thus, false detection can be effectively reduced and the accuracy of blockage identification can be improved.

[0039] The additional aspects and advantages of the present invention will be partially given in the following description, partially become obvious from the following description, or be understood through the practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The above and / or additional aspects and advantages of the present invention will become obvious and easy to understand from the description of the embodiments in conjunction with the following drawings, where:

[0041] Figure 1 Shows a schematic flowchart of the method for monitoring blockage in the crusher feed cavity provided by the embodiment of the present application;

[0042] Figure 2 Shows the preset monitoring area images marked in the images of the feed inlet, the discharge outlet, and the cavity discharge outlet provided by the embodiment of the present application;

[0043] Figure 3 Shows a flowchart for calculating the width ratio of the material in the feed inlet image provided by the embodiment of the present application;

[0044] Figure 4 Shows a model diagram of the material cross-section in the feed inlet provided by the embodiment of the present application;

[0045] Figure 5 Shows a flowchart for calculating the feed cavity level at the cavity discharge outlet provided by the embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0046] The following will clearly describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art belong to the scope of protection of the present application.

[0047] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. are usually of the same type, and the number of objects is not limited. For example, the first object can be one or multiple. In addition, "and / or" in the specification and claims means at least one of the connected objects, and the character " / " generally means that the associated objects before and after are in an "or" relationship.

[0048] In actual mine production, large stones after mining need to be transported to a crusher through a feeder for crushing operations. When the crusher crushes materials, large pieces of raw ore are transported to the feeder inlet of the crusher through a conveyor belt and poured into the crushing chamber for material crushing. The crushed ore with a small particle size is transported to other equipment such as semi-autogenous mills by other conveyor belts for subsequent production. During the entire ore crushing process, factors such as too fast feeding, too large ore particle size, and too much accumulated material in the chamber may all cause blockage of the crushing chamber, thereby damaging the equipment, causing economic losses, and endangering production safety. Therefore, it is extremely necessary to promptly detect and warn of the blockage of the crusher feeder chamber.

[0049] In the existing visual monitoring methods for blockage of the crusher feeder chamber, the method based on deep learning relies on hardware resources and a large number of labeled samples. In actual projects, it is difficult to obtain data, and it is not easy to deploy; the methods based on color segmentation and binocular vision are sensitive to light and color information and have low robustness; the material level detection method based on image difference is prone to false detection of material sputtering and dust falling during the feeding process.

[0050] In summary, there is an urgent need in the industrial field for a fast and effective monitoring method for blockage of the crusher feeder chamber, which can effectively reduce the false alarm rate while ensuring the recognition accuracy.

[0051] The present invention proposes a monitoring method for blockage of the crusher feeder chamber, in combination with Figure 1 , the crusher includes a feed inlet, a discharge outlet, and a chamber connecting the feed inlet and the discharge outlet, and the method includes: step S1, step S2, and step S3.

[0052] S1. Obtain the images of the feeding inlet, discharging outlet, and material chamber in real time, and respectively extract the contours of the materials in the image of the feeding inlet, the contours of the materials in the image of the discharging outlet, and the contours of the materials in the image of the material chamber.

[0053] In this step, first, install monitoring cameras at the observation points of the feeding inlet, discharging outlet, and the discharging opening of the material chamber of the crusher respectively. Through the monitoring cameras, collect the images at the feeding inlet, discharging outlet, and the discharging opening of the material chamber of the crusher, and obtain the images of the feeding inlet, discharging outlet, and the discharging opening of the material chamber in real time, as Figure 2 shown.

[0054] The image of the feeding inlet includes the images of the material and the conveyor belt for transporting the material.

[0055] The image of the discharging outlet includes the images of the material and the conveyor belt for transporting the material.

[0056] The image of the material chamber includes the images of the material and the image of the edge of the discharging opening of the material chamber that is not covered by the material. It should be noted here that the flow direction of the material in the crusher is: feeding inlet → material chamber → discharging opening → discharging outlet, and the discharging opening is located upstream of the discharging outlet.

[0057] In some embodiments, the images collected by the camera need to be labeled before extracting the material contour to obtain a preset monitoring area. Specifically, see Figure 2 , based on the image of the feeding inlet, mark the preset monitoring area in the image and denote it as R in , R in The four vertices of are respectively denoted as A in , B in , C in and D in , where the preset monitoring area R in only includes the images of the material and the conveyor belt for transporting the material; based on the image of the discharging outlet, mark the preset monitoring area in the image and denote it as R out , R out The four vertices of are respectively denoted as A out , B out , C out and D out , where the preset monitoring area R out only includes the images of the material and the conveyor belt for transporting the material; based on the image of the discharging opening of the material chamber, mark the preset monitoring area in the image and denote it as R f , R f The four vertices of are respectively denoted as A f , B f , C f and D f , where the preset monitoring area R fAn image that only contains materials and the edge of the material cavity discharge port that is not covered by materials.

[0058] A method for monitoring material blockage in a crusher feed cavity according to an embodiment of the present application. By obtaining images of the feed inlet, the discharge outlet, and the cavity in real time, and annotating the obtained images to obtain a preset monitoring area, it is beneficial to extract the contour of the material in the image.

[0059] In some embodiments, step S1 may include:

[0060] Using a Gaussian mixture model and an image difference method to extract the contour of the material in the image of the feed inlet; and / or,

[0061] Using a Gaussian mixture model and an image difference method to extract the contour of the material in the image of the discharge outlet; and / or,

[0062] Using a Gaussian mixture model and an image difference method to extract the contour of the material in the image of the cavity.

[0063] In this step, the Gaussian mixture model quantifies things precisely using a Gaussian probability density function (normal distribution curve), decomposes a thing into several models formed based on the Gaussian probability density function (normal distribution curve), and can obtain the background image in the image through the Gaussian mixture model.

[0064] When the monitoring camera is fixed, the change of the background is slow. The conveyor belt in the images of the discharge outlet and the feed inlet belongs to the background. In the image of the cavity, the area except the material part belongs to the background, and the foreground is the moving object. The materials in the images of the discharge and feed inlets and the image of the material cavity discharge port belong to the foreground. By using the Gaussian mixture model to model the background of the images of the discharge outlet, the feed inlet, and the material cavity discharge port, and separating the background image from a given image. Specifically, let the current frame source image be I t , then the method for obtaining the background image I bg corresponding to the current frame is as follows:

[0065] I bg = G(I t-1 , I t-2 , I t-3 ,..., I t-n )

[0066] Among them, n is the number of frames for modeling the Gaussian mixture model. In this embodiment, the background models of the discharge and feed inlets are short-term models, and the number of frames for modeling is n1 = 50, about 2 seconds, which can effectively adapt to the rapid change of the material image; the background model of the cavity is a long-term model, and the number of frames for modeling is n2 = 750, about 30 seconds, which can effectively improve the interference of the model by factors such as material sputtering and dust falling.

[0067] The current frame source image refers to the image of the marked inlet, or the image of the marked outlet, or the image of the marked material chamber.

[0068] The image difference method is to subtract the corresponding pixel values of two images to weaken the similar parts of the images and highlight the changed parts. For example, the difference image can often detect the contour of the moving object, that is, the material contour of the present application, which can also be called the foreground image.

[0069] According to the monitoring method provided by the embodiments of the present application, by performing background modeling on the marked image, the background image and the foreground image of the current image can be obtained, that is, the contour image of the material can be effectively extracted.

[0070] In some embodiments, step S1 may include:

[0071] The crusher feeds the material into the inlet through the first conveyor belt and discharges the material from the outlet through the second conveyor belt. The image of the inlet shows the material on the first conveyor belt and the inlet, and the image of the outlet shows the material on the second conveyor belt and the inlet;

[0072] Use the Gaussian mixture model to obtain the image of the first conveyor belt, and use the image difference method, the image of the first conveyor belt and the image of the inlet to extract the contour of the material in the image of the inlet; and / or,

[0073] Use the Gaussian mixture model to obtain the image of the second conveyor belt, and use the image difference method, the image of the second conveyor belt and the image of the outlet to extract the contour of the material in the image of the outlet; and / or,

[0074] Use the Gaussian mixture model to obtain the image of the material chamber that is not covered by the material, and use the image difference method, the image of the material chamber that is not covered by the material and the image of the material chamber to extract the contour of the material in the image of the material chamber.

[0075] It can be understood that the image of the first conveyor belt is different from the first conveyor belt. Since there is a certain angle between the monitoring camera for obtaining the conveyor belt image and the first conveyor belt, when the first conveyor belt is substantially rectangular, the image of the first conveyor belt may be trapezoidal; similarly, the image of the second conveyor belt is also different from the second conveyor belt.

[0076] The first conveyor belt and the second conveyor belt run at the same speed.

[0077] In this step, the background images of the feeding port, discharging port, and material chamber are obtained through the Gaussian mixture model. Among them, the background image of the feeding port is the image of the first conveyor belt; the background image of the discharging port is the image of the second conveyor belt; the background image of the material chamber is the image of the discharging opening of the material chamber not covered by materials. Finally, the contour of the material is obtained through the image difference method.

[0078] For example, the image of the first conveyor belt is compared with the image of the marked feeding port, and the corresponding pixel values of the two images are subtracted, so as to obtain the contour of the material in the image of the feeding port.

[0079] The method for obtaining the contour of the material in the image of the discharging port and the method for obtaining the contour of the material in the image of the discharging opening of the material chamber are the same as the method for obtaining the contour of the material in the image of the feeding port, and will not be elaborated here.

[0080] In some embodiments, after step S1, it further includes:

[0081] Calculate the value of m in / M in of the nth frame of the feeding port image, where m in is the area of the material image in the nth frame of the feeding port image, and M in is the area of the nth frame of the feeding port image;

[0082] If the value of m in / M in of the nth frame of the feeding port image is greater than the area threshold T a , and the difference between the value of m in / M in of the nth frame of the feeding port image and the value of m in / M in of the (n - 1)th frame of the feeding port image is greater than the mutation threshold T s , then repeat step S1.

[0083] Due to the lighting requirements at the production site, the lighting will change suddenly during the day-night alternation, resulting in misdetection of the moving foreground. In this step, based on the image of the marked feeding port, for the contour of the material in each frame of the feeding port image, it is necessary to calculate the area ratio m in / M in of the material contour in the preset monitoring area R in , where m in is the area of the material image in the nth frame of the feeding port image, and M in is the area of the nth frame of the feeding port image; if the area ratio m in / M in of the material contour in the current frame of the feeding port image in the preset monitoring area R in is greater than the area threshold T a, and the ratio of the area of the material contour in the image of the feeding port in the previous frame to the area m in the preset monitoring area R in in in / M in has a difference greater than the mutation threshold T s , it is considered that a light mutation occurs in the current frame, and the background model of the image of the feeding port needs to be re-initialized and modeled. In this embodiment, the area threshold T a is 0.85, and the mutation threshold T s is 0.5.

[0084] In some embodiments, after step S1, it further includes:

[0085] Calculate the value of m out / M out for the image of the discharge port in the nth frame, where m out is the area of the image of the material in the image of the discharge port in the nth frame, and M out is the area of the image of the discharge port in the nth frame;

[0086] If the value of m out / M out for the image of the discharge port in the nth frame is greater than the area threshold T′ a , and the difference between the value of m out / M out for the image of the discharge port in the nth frame and the value of m out / M out for the image of the discharge port in the (n - 1)th frame is greater than the mutation threshold T′ s , then repeat step S1.

[0087] In this step, based on the annotated image of the discharge port, for the contour of the material in the image of the discharge port in each frame, it is necessary to calculate the ratio of the area of the material contour in the preset monitoring area R out in out m out / M out , where m out is the area of the image of the material in the image of the discharge port in the nth frame, and M out is the area of the image of the discharge port in the nth frame; if the ratio of the area of the material contour in the image of the current discharge port in the preset monitoring area R out m out / M a is greater than the area threshold T′ out , and the difference from the ratio of the area of the material contour in the image of the previous discharge port in the preset monitoring area R out m out / M s is greater than the mutation threshold T′, it is considered that a light mutation occurs in the current frame, and the background model of the image of the discharge port needs to be re-initialized and modeled. In this embodiment, the area threshold T′a is 0.85, and the mutation threshold T' s is 0.5.

[0088] In some embodiments, after step S1, the following is further included:

[0089] Calculate the value of m f / M f of the nth frame of the cavity image, where m f is the area of the image of the material in the nth frame of the cavity image, and M f is the area of the nth frame of the cavity image;

[0090] If the value of m f / M f of the nth frame of the cavity image is greater than the area threshold T″ a , and the difference between the value of m f / M f of the nth frame of the cavity image and the value of m f / M f of the (n - 1)th frame of the cavity image is greater than the mutation threshold T″ s , then repeat step S1.

[0091] In this step, based on the image of the annotated cavity, for the contour of the material in each frame of the cavity image, it is necessary to calculate the area ratio m f / M f in the preset monitoring area R f , where m f is the area of the image of the material in the nth frame of the cavity image, and M f is the area of the nth frame of the cavity image; if the area ratio m f / M f of the contour area of the material in the current frame of the cavity image in the preset monitoring area R f is greater than the area threshold T″ a , and the difference from the area ratio m f / M f of the contour area of the material in the image of the previous frame of the cavity in the preset monitoring area R f is greater than the mutation threshold T″ s , it is considered that a light mutation occurs in the current frame, and the background model of the cavity image needs to be re - initialized and modeled. In this embodiment, the area threshold T″ a is 0.85, and the mutation threshold T″ s is 0.5.

[0092] According to the detection method provided by the present application, the background update strategy designed for interference factors such as light mutation and dust deposition can effectively improve the robustness of the material contour detection algorithm.

[0093] S2. Calculate the unit flow rate S of the material at the inlet according to the contour of the material in the image of the inlet in , calculate the unit flow rate S of the material at the outlet according to the contour of the material in the image of the outlet out , ΔS = α·S in -S out , is the flow rate of the material at the inlet per unit time, is the flow rate of the material at the outlet per unit time. Calculate the material level d in the material chamber according to the contour of the material in the image of the material chamber discharge port

[0094] In this step, the unit flow rate refers to the volume of the material passing through per unit time

[0095] Since the average particle size of the material fed into the crusher (i.e., the average particle size at the inlet) is usually larger than the particle size of the crushed material at the outlet (i.e., the average particle size of the material at the outlet), it is necessary to statistically calculate the density ratio parameter of the same material before and after crushing to avoid the volume difference of the material stack caused by the particle size difference, so as to affect the estimation of the difference between the inlet and outlet material quantities. Specifically, when the crusher is operating normally, the inlet and outlet material quantities should be the same. At this time, the average unit flow rates of the inlet and outlet materials within a certain period of time can be estimated respectively and Calculate the inlet and outlet density ratio parameter

[0096] After obtaining the inlet and outlet density ratio parameter through statistics, the on-line measurement of the difference between the inlet and outlet material quantities of the material can be carried out. Specifically, first obtain the inlet unit flow rate S in and the outlet unit flow rate S out , then the difference between the inlet and outlet material quantities ΔS = α·S in -S out .

[0097] In some embodiments, calculating the inlet unit flow rate S in includes: taking a vertex in the image of the inlet as the origin, the width direction of the first conveyor belt as the abscissa X-axis, and the direction perpendicular to the X-axis as the ordinate Y-axis to establish a coordinate system. Divide the contour of the material in the image of the inlet along the Y-axis. The absolute value of the coordinate of the material in the image of the inlet close to the origin on the Y-axis is the minimum ordinate y min , the absolute value of the coordinate far from the origin on the Y-axis is the maximum ordinate y max , calculate the average proportion P of the material width in the ordinate interval [y min , y max : in :

[0098]

[0099] pi = h in，i / H in，i ,h in，i is the i-th position in the outline of the material in the image of the feeding port, the width of the material, H in，i is the width of the first conveyor belt corresponding to the i-th position in the outline of the material in the image of the feeding port;

[0100] w′ in = P in · w in ,w in is the width of the first conveyor belt in the image of the feeding port. Calculate S according to the width w′ of the outline of the material in the image of the feeding port in Calculate S in .

[0101] In this step, first, it is necessary to determine the minimum ordinate y in and the maximum ordinate y min of the conveyor belt according to the preset monitoring area R of the feeding port max . Specifically, according to the preset monitoring area R of the feeding port in , taking the line where A in B in is located as the X-axis and the line perpendicular to A in B in as the Y-axis, and taking the intersection of the X-axis and the Y-axis as the origin to construct a coordinate system. See Figure 3 . The coordinate of the point on the Y-axis close to the A in point is the minimum ordinate y min , and the coordinate of the point far from the A in point is the maximum ordinate y max . Traverse the ordinate interval [y min , y max and calculate the ratio p n of the width of the first conveyor belt corresponding to the material width at each ordinate y i in the current row, where y n belongs to [y min , y max . The calculation method of p i is as follows:

[0102]

[0103]

[0104] Among them, x left and x right respectively represent the left and right abscissa edges of the first conveyor belt when the ordinate is y n ; f(x, y n ) is the foreground state function. When the pixel pixel(x, yn ) When it is greater than 0, that is, pixel > 0, when it is the foreground of the material, the value is 1, and otherwise it is 0.

[0105] It can be understood that the "width" in the embodiments of the present invention refers to the dimension in the X direction in the coordinate system.

[0106] Then, calculate the average proportion P min , y max of the material width within the ordinate interval [y in , and the calculation method is as follows:

[0107]

[0108] where n is [y min , y max , and i belongs to [y min , y max .

[0109] Combined with the average proportion P in of the material width and the material cross-section model to estimate the unit flow rate S in of the material. Specifically, referring to Figure 4 , through the cross-sectional view of the material passing through the inlet, a material cross-section model is constructed. The overall cross-section model includes the cross-section of the first conveyor belt, the lower stockpile cross-section, and the upper stockpile cross-section. Among them, the cross-section of the first conveyor belt is approximately a parabola. The parameters required for the parabola function include: the belt opening width w in and the height h in,i can both be obtained through physical measurement; the lower stockpile cross-section is a sub-region of the cross-section of the first conveyor belt, and its width w' in = P in · w in , and its height h' in，i can be calculated through the belt parabola equation, and its cross-sectional area S bottom can be calculated through parabola integration; the upper stockpile cross-section is a triangular region, with the same width and height as the lower stockpile cross-section, and its cross-sectional area S up can be obtained through triangle area calculation.

[0110] Thus, the cross-sectional area S in of the material at the inlet can be obtained, where S in = S bottom + S up , and this cross-sectional area can be regarded as the unit flow rate S in of the material.

[0111] In some embodiments, calculate the unit flow rate S outIncluding: taking a vertex in the image of the discharge port as the origin, the width direction of the second conveyor belt as the abscissa X-axis, and the direction perpendicular to the X-axis as the ordinate Y-axis to establish a coordinate system, dividing the contour of the material in the image of the discharge port along the Y-axis, and the absolute value of the coordinate of the material in the image of the discharge port close to the origin on the Y-axis is the minimum ordinate y′ min and the absolute value of the coordinate of the material in the image of the discharge port far from the origin on the Y-axis is the maximum ordinate y′ max Calculating the average proportion P of the material width in the ordinate interval [y′ min , y′ max : out

[0112]

[0113] p′ i =h out,i / H out，i where h out，i is the width of the material at the i-th position in the contour of the material in the image of the discharge port, and H out，i is the width of the conveyor belt corresponding to the i-th position in the contour of the material in the image of the discharge port;

[0114] w′ out =P out ·w out where w out is the width of the first conveyor belt in the image of the inlet port. According to the width w′ out of the contour of the material in the image of the inlet port, calculate S out .

[0115] In this step, first, it is necessary to determine the minimum ordinate y′ out and the maximum ordinate y′ min of the conveyor belt according to the preset monitoring area R max of the discharge port. Specifically, according to the preset monitoring area R out of the discharge port, taking the straight line where A out B out is located as the X-axis, the straight line perpendicular to A out B out as the Y-axis, and the intersection point of the X-axis and the Y-axis as the origin to construct a coordinate system. The coordinate of the point on the Y-axis close to the A out point is the minimum ordinate y′ min , and the coordinate of the point far from the A out point is the maximum ordinate y′ max . Traverse the ordinate interval [y′ min , y′ max and calculate the proportion p′ n of the material width at each ordinate y′ i ​, where y' n belongs to [y' min , y' max , and the calculation method of p' i is as follows:

[0116]

[0117]

[0118] Among them, x' left and x' right respectively represent the left and right horizontal coordinate edges of the second conveyor belt when the vertical coordinate is y' n ; f(x', y' n ) is the foreground state function. When the pixel pixel(x', y' n ) > 0, that is, when the pixel (pixel) > 0 and it is the foreground of the material, the value is 1, and otherwise it is 0.

[0119] Then, calculate the average proportion P out of the material width within the vertical coordinate interval [y' min , y' max , and the calculation method is as follows:

[0120]

[0121] Among them, n is [y' min , y' max , and i belongs to [y' min , y' max .

[0122] Combined with the average proportion P out of the material width and the material cross-section model to estimate the unit flow rate S out of the material. Specifically, through the cross-sectional view of the material at the discharge port, a material cross-section model is constructed. The overall cross-section model includes the cross-section of the second conveyor belt, the lower stockpiling cross-section, and the upper stockpiling cross-section. Among them, the cross-section of the second conveyor belt is approximately a parabola, and the parameters required for the parabola function include: the belt opening width w out and the height h out,i can both be obtained through physical measurement; the lower stockpiling cross-section is a sub-region of the cross-section of the second conveyor belt, and its width w' out = P out ·w out , and its height h' out，i can be calculated through the belt parabola equation, and its cross-sectional area S' bottom can be calculated through parabola integration; the upper stockpiling cross-section is a triangular region with the same width and height as the lower stockpiling cross-section, and its cross-sectional area S' can be obtained through triangle area calculation.up 。

[0123] Thus, the cross-sectional area S of the material at the discharge port can be obtained out , where S out = S′ bottom + S′ up , and this cross-sectional area can be regarded as the unit flow rate S of the material at the discharge port out 。

[0124] In some embodiments, according to the material contour at the material discharge opening of the material chamber, the material level d of the feeding chamber is calculated

[0125] In this step, referring to Figure 5 , based on the image of the material discharge opening of the material chamber, a preset monitoring area is marked in the image and denoted as R f , R f The four vertices of are respectively denoted as A f , B f , C f and D f . According to the preset monitoring area R f , the material contour in the image of the material discharge opening of the material chamber is obtained. Taking the line connecting A f B f as the marking baseline, the pixel farthest from the marking baseline A f B f is obtained according to pixel statistics, and the distance between this pixel and the baseline A f B f is the material level d of the feeding chamber

[0126] The baseline refers to the position close to the edge of the material discharge opening of the material chamber

[0127] Furthermore, for the material level d and the baseline A f B f , it is necessary to satisfy that for each row of material foreground pixels from A f B f to the material level d, the foreground ratio is greater than the third preset value t h 。

[0128] It should be noted that an image is actually composed of pixels. A pixel can be regarded as a rectangle, and each row of material foreground pixels can be regarded as each row being composed of single pixel blocks juxtaposed; the foreground ratio refers to the ratio of the total number of pixel blocks of the foreground image in a row to the total number of pixel blocks in that row. Here, the total pixels include the material pixels and background pixels in that row

[0129] A monitoring method for material blockage in the feeding chamber of a crusher provided by an embodiment of the present application performs material detection and segmentation based on the background modeling method, and designs a background update strategy for interference factors such as sudden light changes and dust deposition, which can effectively improve the robustness of the material detection algorithm; accurately estimates the material flow rate through the material cross-section model, and uses the inlet and outlet density ratio parameter to correct the volume change before and after the material pile is broken. Compared with the prior art, this strategy greatly improves the accuracy of estimating the inlet and outlet material quantities of the crusher, and at the same time effectively reduces the interference of the volume change caused by material crushing on the statistics of the difference between the inlet and outlet material quantities.

[0130] S3. Calculate the sum ΔS of ΔS within a preset time sum , if ΔS sum is greater than the first preset value and d is greater than the second preset value, then give an alarm.

[0131] In this step, calculate the sum ΔS of ΔS within a preset time sum , by comparing the calculated difference ΔS sum between the inlet and outlet material quantities with the first preset value, and comparing the calculated material level d in the feeding chamber with the second preset value. If ΔS sum is greater than the first preset value and d is greater than the second preset value, it indicates that the feeding chamber of the crusher is blocked, and at this time, a blockage alarm needs to be given. Specifically, let the current frame be the t-th frame and the number of frames for blockage analysis be M. Then, within the time period from t - m to t frames (i.e., within the preset time), calculate the total difference in the inlet and outlet material quantities of the crusher:

[0132]

[0133] where m belongs to [0, M] representing the frame index, and S t-m represents the difference in the inlet and outlet material quantities of the historical frame with a distance of m frames from the current frame.

[0134] Set the first preset value for the difference in material quantities according to ΔS sum . When ΔS sum is greater than the first preset value, it indicates that the inlet and outlet material quantities are abnormal, and at this time, blockage may occur, entering the material quantity warning state;

[0135] For the measurement results of the material level in the feeding chamber for M consecutive frames, if d is greater than the preset second preset value in all cases, blockage may occur, entering the material level warning state;

[0136] If in the above-mentioned M consecutive frames of images, if ΔS is greater than the first preset value and d is greater than the second preset value, it indicates that the crusher is blocked, and at this time, a blockage alarm needs to be given.

[0137] According to the monitoring method for material blockage in the crusher feeding chamber provided by the embodiments of the present application, material quantity difference early warning and material level early warning analysis are combined on the basis of timing analysis. Compared with the prior art, it can effectively reduce false detection and improve the accuracy of material blockage identification.

[0138] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application and are not intended to limit them. Although the present application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A monitoring method for material blockage in the feeding chamber of a crusher, the crusher comprising a feeding port, a discharging port, and a material chamber connecting the feeding port and the discharging port, characterized in that, The method includes: S1. Obtain the images of the inlet, the outlet, and the material chamber in real time, and respectively extract the contours of the material in the image of the inlet, the contours of the material in the image of the outlet, and the contours of the material in the image of the material chamber; S2. Calculate the unit flow rate S of the material at the inlet according to the contour of the material in the image of the inlet in , calculate the unit flow rate S of the material at the outlet according to the contour of the material in the image of the outlet out , ΔS = α·S in -S out , is the flow rate of the material at the inlet per unit time, is the flow rate of the material at the outlet per unit time, and calculate the material level d of the material in the cavity according to the contour of the material in the image of the cavity; S3. Calculate the sum of ΔS, denoted as ΔS, within a preset time sum , if ΔS sum is greater than the first preset value and d is greater than the second preset value, then an alarm is issued.

2. The method according to claim 1, characterized in that, In step S1, Use the Gaussian mixture model and the image difference method to extract the contour of the material in the image of the inlet; and / or Use the Gaussian mixture model and the image difference method to extract the contour of the material in the image of the outlet.

3. The method according to claim 1, characterized in that In step S1, use the Gaussian mixture model and the image difference method to extract the contour of the material in the image of the material chamber.

4. The method according to claim 2 or 3, characterized in that, In step S2, the crusher feeds the material into the inlet through the first conveyor belt and conveys the material out of the outlet through the second conveyor belt. The image of the inlet shows the material on the first conveyor belt and the inlet, and the image of the outlet shows the material on the second conveyor belt and the inlet; Use the Gaussian mixture model to obtain the image of the first conveyor belt, and use the image difference method and based on the image of the first conveyor belt and the image of the inlet to extract the contour of the material in the image of the inlet; and / or Use the Gaussian mixture model to obtain the image of the second conveyor belt, and use the image difference method and based on the image of the second conveyor belt and the image of the outlet to extract the contour of the material in the image of the outlet; and / or Use the Gaussian mixture model to obtain the image of the un-covered part of the material chamber's discharge port by the material, and use the image difference method and based on the image of the un-covered part in the material chamber and the image of the material chamber to extract the contour of the material in the image of the material chamber.

5. The method according to claim 4, wherein Taking a vertex in the inlet image as the origin, the width direction of the first conveyor belt as the abscissa X-axis, and the direction perpendicular to the X-axis as the ordinate Y-axis to establish a coordinate system, the contour of the material in the inlet image is segmented along the Y-axis. The absolute value of the coordinate of the material closest to the origin on the Y-axis in the inlet image is the minimum ordinate y min , and the absolute value of the coordinate farthest from the origin on the Y-axis is the maximum ordinate y max , calculate the average proportion P min of the material width within the ordinate interval [y max , y in : p i = h in，i / H in，i ,where h in，i is the width of the material at the i-th position in the contour of the material in the image of the inlet, and H in.i is the width of the first conveyor belt corresponding to the i-th position in the contour of the material in the image of the inlet; w′ in = P in ·w in ,w in is the width of the first conveyor belt in the image of the feed inlet, and S is calculated according to the width w′ of the outline of the material in the image of the feed inlet in Calculate S in .

6. The method according to claim 4, wherein Taking a vertex in the image of the discharge port as the origin, the width direction of the second conveyor belt as the abscissa X-axis, and the direction perpendicular to the X-axis as the ordinate Y-axis to establish a coordinate system, dividing the contour of the material in the image of the discharge port along the Y-axis, and the absolute value of the coordinate of the material closest to the origin on the Y-axis in the image of the discharge port is the minimum ordinate y'. min The absolute value of the coordinate farthest from the origin on the Y-axis is the maximum ordinate y'. max Calculate the average proportion P of the material width in the ordinate interval [y' min , y' max : out :[[]]END]] p′ i = h out，i / H out，i , h out.i is the width of the material at the i-th position in the contour of the material in the image of the discharge port, H out.i is the width of the conveyor belt corresponding to the i-th position in the contour of the material in the image of the discharge port; w' out = P out · w out ,w out is the width of the second conveyor belt in the image of the discharge port, and S is calculated according to the width w' of the contour of the material in the image of the discharge port out Calculate S out .

7. The method according to claim 4, wherein Between step S1 and S2, it further includes: Calculate the value of m of the nth frame inlet image in / M in , where m in is the area of the material image in the nth frame inlet image, and M in is the area of the nth frame inlet image; If the value of m in the n-th frame of the feed inlet image in / M in is greater than the area threshold Ta, and the difference between the value of m in the n-th frame of the feed inlet image in / M in and the value of m in the (n - 1)-th frame of the feed inlet image in / M in is greater than the mutation threshold T s , then repeat step S1.

8. The method according to claim 4, characterized in that, Between step S1 and S2, it further includes: Calculate the value of m for the nth frame discharge port image out / M out , where m out is the area of the material image in the nth frame discharge port image, and M out is the area of the nth frame discharge port image; If the value of m out / M out in the n-th frame discharge port image is greater than the area threshold T′ a , and the difference between the value of m out / M out in the n-th frame discharge port image and the value of m out / M out in the (n - 1)-th frame discharge port image is greater than the mutation threshold T′ s , then repeat step S1.

9. The method according to claim 4, wherein Between step S1 and S2, it further includes: Calculate the value of m of the nth frame cavity image f / M f , where m f is the area of the material image in the nth frame cavity image, and M f is the area of the nth frame cavity image; If the value of m f / M f in the n-th frame cavity image is greater than the area threshold T″ a , and the difference between the value of m f / M f in the n-th frame cavity image and the value of m f / M f in the (n - 1)-th frame cavity image is greater than the mutation threshold T″ s , then repeat step S1.

10. The method according to claim 1, wherein Step S3 further includes: If ΔS is less than or equal to the first preset value and d is greater than the second preset value, then issue a warning; If ΔS is greater than the first preset value and d is less than or equal to the second preset value, then issue a warning.

Images