阅读说明：本技术 通过监控视频实现球磨机设备工作状态检测的方法及系统 (Method and system for realizing detection of working state of ball mill equipment through monitoring video ) 是由 叶中付 于 2020-05-18 设计创作，主要内容包括：本发明公开了一种通过监控视频实现球磨机设备工作状态检测的方法及系统,方法包括：在球磨机筒体外表面位置设置能同步转动的标记单元；在球磨机筒体外部设置摄像机,清晰录制标记单元随筒体绕其轴心同步转动的标记转动视频,利用图像识别处理实时识别出标记转动视频中的标记单元,并转换为周期性相关系数曲线；检测出单位时间内相关系数曲线上升沿的总数,根据上升沿总数计算出球磨机筒体的转速、转动角速度和转动周期；根据工作时球磨机的转速、转动角速度和转动周期的任一参数的实际值与正常值的大小关系,确定球磨机当前工作状态,并在其为非正常工作状态时进行相应处理。能实现对球磨机工作状态自动、实时和准确的监测。(The invention discloses a method and a system for realizing the detection of the working state of ball mill equipment through a monitoring video, wherein the method comprises the following steps: a marking unit capable of synchronously rotating is arranged on the outer surface of the ball mill cylinder; arranging a camera outside a ball mill cylinder, clearly recording a marking rotation video of a marking unit along with synchronous rotation of the cylinder around the axis of the cylinder, identifying the marking unit in the marking rotation video in real time by utilizing image identification processing, and converting the marking unit into a periodic correlation coefficient curve; detecting the total number of rising edges of the relation number curve in unit time, and calculating the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder according to the total number of the rising edges; according to the relation between the actual value and the normal value of any parameter of the rotating speed, the rotating angular speed and the rotating period of the ball mill during working, the current working state of the ball mill is determined, and corresponding processing is carried out when the ball mill is in an abnormal working state. The automatic, real-time and accurate monitoring of the working state of the ball mill can be realized.)

1. A method for realizing the detection of the working state of ball mill equipment through monitoring videos is characterized by comprising the following steps:

the outer surface of the cylinder body of the ball mill is provided with a marking unit which is convenient for image recognition processing, and the marking unit can synchronously rotate around the axis of the cylinder body along with the outer surface of the cylinder body;

arranging a camera on the periphery of the ball mill cylinder, and clearly recording a marking operation video of the marking unit rotating synchronously with the ball mill cylinder through the camera;

identifying a marking unit in the marking operation video through the image identification processing, converting the identified marking unit into a correlation coefficient in real time, and outputting the correlation coefficient as a periodic correlation coefficient curve;

detecting to obtain the total number of the rising edges of the correlation coefficient curve in unit time, and calculating to obtain the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder during working according to the total number of the rising edges of the correlation coefficient curve in unit time;

and determining the current working state of the detected ball mill according to the calculated size relationship between the actual value and the normal value of any parameter in the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder during working, and performing corresponding processing when the ball mill is in the abnormal working state.

2. A method for detecting the working condition of a ball mill equipment through monitoring video according to claim 1, characterized in that, in the method, the marking unit which is arranged on the outer surface of the ball mill cylinder and is convenient for the image recognition processing to recognize comprises the following steps:

arranging a marking unit on the arc-shaped outer surface of the ball mill cylinder;

or, a marking unit is arranged on the surface of any side of the ball mill cylinder body at a position deviating from the axis of the ball mill cylinder body.

3. A method for detecting the working condition of a ball mill equipment through monitoring video according to claim 1 or 2, characterized in that, in the method, the cameras arranged on the periphery of the ball mill cylinder are:

and arranging a camera at a position where the marking unit can record the clear video synchronously rotating along with the cylinder on the periphery of the cylinder of the ball mill.

4. The method for detecting the working state of the ball mill equipment through the monitoring video according to the claim 1 or 2, characterized in that in the method, the marking units in the marking operation video are identified through the image identification processing, the identified marking units are converted into the correlation coefficients in real time, and the correlation coefficients are output as a periodic correlation coefficient curve;

in the method, a marking unit in the marking operation video is identified through the image identification processing, the identified marking unit is converted into a correlation coefficient in real time, and the correlation coefficient is output as a periodic correlation coefficient curve:

presetting matching processing related to the marking unit pixel in the image identification processing, wherein the matching processing can output a correlation coefficient after the marking unit pixel is identified;

according to the correlation coefficient, calculating a correlation coefficient curve through a correlation coefficient curve formula x (t) ═ s (t) + n (t), and outputting the correlation coefficient curve, wherein s (t) is the correlation coefficient of the marking unit pixel in the fixed detection window of the video frame at the time t and the matching processing preset in the image identification processing; n (t) is noise generated by the correlation operation between the non-marked unit pixels in the fixed detection window of the video frame at the time t and the matching process preset in the image identification process, and the n (t) is ignored.

5. The method for detecting the working state of the ball mill equipment through the surveillance video according to claim 1 or 2, wherein the total number of the rising edges of the correlation coefficient curve in unit time obtained through detection is as follows:

at a set sampling period T_sSampling and outputting the waveform of the correlation coefficient curve, and obtaining the total number N of the rising edges of the correlation coefficient curve in unit time T from a sampling result;

the rotating speed of the ball mill cylinderWherein T represents a unit time, and N represents the total number of rising edges of the correlation coefficient curve in the unit time T;

the rotation angular velocity of the ball mill cylinderWherein N represents the rotation speed of the cylinder, T represents unit time, and N represents the rotation speed of the cylinder in the unit time TThe total number of rising edges of the correlation coefficient curve;

the rotation period of the ball mill cylinderWherein N represents the rotation speed, T represents the unit time, and N represents the total number of the rising edges of the correlation coefficient curve in the unit time T.

6. A method for realizing detection of an operating state of a ball mill equipment through monitoring videos according to claim 5, characterized in that in the method, the rising edge of the correlation coefficient curve is determined by the following method:

if the difference value of the correlation coefficient curve waveform of the ith sampling point and the last sampling point is calculated_iNot less than a predetermined threshold value₀And then confirming that the sampling point is the rising edge of the correlation coefficient curve, the serial number of the sampling point corresponding to the rising edge of the correlation coefficient curve is i, and accumulating and counting the number of the rising edges of the correlation coefficient curve to obtain the total number N of the rising edges of the correlation coefficient curve in unit time T.

7. The method for detecting the working state of the ball mill equipment through the monitoring video according to claim 1 or 2, wherein in the method, according to the calculated magnitude relation between the actual value and the normal value of any one parameter of the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder during working, the current working state of the detected ball mill is determined, and when the ball mill is in the abnormal working state, the corresponding processing comprises the following steps:

(1) determining the current working state of the detected ball mill according to the size relation between the actual value and the normal value of the rotating speed of the ball mill cylinder, and performing corresponding processing when the ball mill is in an abnormal working state, wherein the method comprises the following steps:

the rotation speed range of the cylinder body in the normal rotation speed working state of the ball mill to be detected is set as n in advance₁≤n≤n₂Dividing the rotating speed of the ball mill cylinder into three sets according to different working states:

X_{high speed}＝{n|n＞n₂}

X_{Is normal}＝{n|n₁≤n≤n₂}

X_{Low speed}＝{n|n＜n₁}；

Calculating the actually measured rotating speed n of the ball mill cylinder in the production process, and if the rotating speed n ∈ X is the rotating speed n_{Is normal}Determining that the ball mill is in a normal rotating speed working state, maintaining the current rotating speed, and if the rotating speed is n ∈ X_{High speed}Determining that the ball mill is in a high-rotation-speed working state, and giving out an alarm and/or reducing the rotation speed, if the rotation speed is n ∈ X_{Low speed}Determining that the ball mill is in a low-rotation-speed working state, and performing processing of giving an alarm and/or increasing the rotation speed;

(2) determining the current working state of the detected ball mill according to the size relation between the actual value and the normal value of the rotation angular speed of the ball mill cylinder, and performing corresponding processing when the ball mill is in an abnormal working state, wherein the method comprises the following steps:

the rotation angular speed range of the cylinder body in the normal rotation speed working state of the ball mill to be detected is set to omega in advance₁≤ω≤ω₂Dividing the rotation angular speed of the ball mill cylinder into three sets according to different working states:

Y_{high speed}＝{ω|ω＞ω₂}

Y_{Is normal}＝{ω|ω₁≤ω≤ω₂}

Y_{Low speed}＝{ω|ω＜ω₁}；

Calculating the actually measured rotation angular speed omega of the cylinder body of the ball mill in the production process, and if the rotation angular speed omega ∈ Y_{Is normal}Determining that the ball mill is in normal rotation speed working state, maintaining current rotation speed, and if the rotation angular speed is omega ∈ Y_{High speed}Determining that the ball mill is in high-rotation-speed working state, making alarm and/or reducing rotation speed, if its rotation angular speed is omega ∈ Y_{Low speed}Determining that the ball mill is in a low-rotation-speed working state, and performing processing of giving an alarm and/or increasing the rotation speed;

(3) determining the current working state of the detected ball mill according to the size relation between the actual value and the normal value of the rotation period of the ball mill cylinder, and performing corresponding processing when the ball mill is in an abnormal working state, wherein the method comprises the following steps:

the rotation period range of the cylinder body in the normal rotation speed working state of the ball mill to be detected is preset to be T₁≤T_ω≤T₂Dividing the rotation period of the ball mill cylinder into three sets according to different working states:

Z_{high speed}＝{T_ω|T_ω＜T₁}

Z_{Is normal}＝{T_ω|T₁≤T_ω≤T₂}

Z_{Low speed}＝{T_ω|T_ω＞T₂}；

Calculating the actually measured rotation period T of the ball mill cylinder in the production process_ωIf the period of rotation T_ω∈Z_{Is normal}Determining that the ball mill is in a normal rotating speed working state, and maintaining the current rotating speed; if the period of rotation T_ω∈Z_{High speed}Determining that the ball mill is in a high-rotation-speed working state, and performing processing of giving an alarm and/or reducing the rotation speed; if the period of rotation T_ω∈Z_{Low speed}And determining that the ball mill is in a low-rotation-speed working state, and performing processing of giving an alarm and/or increasing the rotation speed.

8. A method for detecting the operation status of a ball mill apparatus according to claim 7, wherein the rotation speed range of the barrel body for detecting the normal rotation speed operation status of the ball mill is set to n in advance₁≤n≤n₂The rotation speed range of the cylinder body in the normal rotation speed working state of the ball mill to be detected is set as n according to prior knowledge accumulated in the actual production process₁≤n≤n₂；

The rotation angular speed range of the cylinder body in the normal rotation speed working state of the ball mill to be detected is set to omega in advance₁≤ω≤ω₂Setting the rotation angular speed range of the cylinder body in the normal rotation speed working state of the ball mill to be detected as omega according to the prior knowledge accumulated in the actual production process₁≤ω≤ω₂；

Setting the range of the rotation period of the cylinder in the normal rotation speed working state of the ball mill to be detected as T in advance₁≤T_ω≤T₂Setting the rotation period range of the cylinder body in the normal rotation speed working state of the ball mill to be detected as T according to the prior knowledge accumulated in the actual production process₁≤T_ω≤T₂。

9. A system for detecting the working condition of a ball mill device through monitoring video, which is used for realizing the method of any one of claims 1 to 8, and comprises the following steps:

the device comprises a marking unit, a camera, an image recognition device, a sampling processing device, a detection processing device and an abnormal state processing device; wherein the content of the first and second substances,

the marking unit is arranged on the outer surface of the cylinder of the detected ball mill and can synchronously rotate around the axis of the cylinder along with the outer surface of the cylinder;

the camera is arranged on the periphery of the ball mill cylinder and can clearly record a marking operation video of the marking unit rotating synchronously along with the ball mill cylinder;

the image recognition device is respectively in communication connection with the output end of the camera and the sampling processing device, recognizes the motion track of a marking unit in the marking operation video, converts the recognized motion track into a correlation coefficient in real time, and outputs a periodic correlation coefficient curve to the sampling processing device according to the correlation coefficient;

the sampling processing device can detect the correlation coefficient curve, detect the total number of the rising edges of the correlation coefficient curve in unit time, and calculate the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder during working according to the total number of the rising edges of the correlation coefficient curve in unit time;

the detection processing device is in communication connection with the detection processing device and can determine the current working state of the detected ball mill according to the rotating speed, the rotating angular speed and the size relation between the actual value and the normal value of any parameter in the rotating period of the ball mill cylinder during working;

and the abnormal state processing device can perform corresponding processing when the ball mill is in the abnormal working state according to the current working state of the detected ball mill determined by the detection processing device.

10. The system for realizing the detection of the working state of the ball mill equipment through the monitoring video according to the claim 9, characterized in that a matching module related to the marking unit is arranged in the image recognition device, and can output a correlation coefficient after recognizing the marking unit in the recognition mark rotation video, and output a periodic correlation coefficient curve according to the correlation coefficient;

the sampling processing apparatus includes: the device comprises a sampling device, a counting device and a signal processing device; wherein the content of the first and second substances,

the sampling device is respectively in communication connection with the counting device and the signal processing device and can be used for sampling in a preset sampling period T_sSampling the correlation coefficient curve, and outputting the amplitude of the correlation coefficient curve at each sampling point;

the counting device can calculate the total number of the rising edges of the correlation coefficient curve in unit time T according to the amplitude of the correlation coefficient curve at each sampling point output after sampling by the sampling device;

the signal processing device can obtain the total number of the rising edges of the correlation coefficient curve in the unit time T through the first formula according to the calculation of the counting deviceCalculating the rotating speed N of the ball mill cylinder, wherein in the first formula, T represents unit time, and N represents the total number of rising edges of a correlation coefficient curve in unit time; by a second formulaCalculating to obtain the rotation angular velocity omega of the ball mill, wherein in the second formula, N represents the rotation speed of the ball mill cylinder, T represents unit time, and N represents the total number of rising edges of a correlation coefficient curve in unit time; and by a third formulaCalculating to obtain the rotation period T of the ball mill_ωIn the third formula, N represents the rotating speed of the ball mill cylinder, T represents unit time, and N represents the total number of rising edges of the correlation coefficient curve in unit time;

and an alarm circuit and/or a self-control circuit are arranged in the abnormal state processing device, the alarm circuit can give an alarm for an abnormal working state, and the self-control circuit can adjust the rotating speed of the ball mill in the abnormal working state.

Technical Field

The invention relates to the field of detection of the working state of a ball mill, in particular to a method for realizing the detection of the working state of ball mill equipment through a monitoring video.

Background

In the existing ball mill equipment working ringIn order to improve the production efficiency and reduce the cost of replacing the lining plate on the inner wall of the ball mill, the production process needs to be controlled by adjusting the rotating speed of the ball mill according to the working state of the ball mill. The working state of the ball mill can be divided into low rotating speed (angular speed omega < omega)₁) State, Normal rotational speed (angular velocity ω)₁≤ω≤ω₂) State and high rotational speed (angular speed ω > ω₂) Status. Fig. 1 to 3 are schematic views of three operating states of the ball mill, including the start and landing points of the grinding media balls 10 (hereinafter referred to as media balls) and the material line 30 formed by the sliding material 20. Based on physical knowledge, on the premise of a certain material volume in the ball mill, the positions of a throwing point, a dropping point and a material line of a medium ball are all functions of the rotating speed omega of the ball mill. In actual production, enterprises expect that the ball mill is always in a normal rotating speed state, in the normal state, the media ball has a high initial throwing point, the impact is generated below the vicinity of a material line, namely the impact is mainly performed on the material, the kinetic energy is high, and the material grinding effect is best. When the ball mill is in a low-rotating-speed state, the media ball has a low throwing point, impacts the materials at a far position below the material line and even rolls in the materials, the materials are not sufficiently ground, and the production efficiency is low. When the ball mill is in a high-rotating-speed state, the media balls are higher in throwing point, and impact occurs above the material line, namely the grinding media balls mainly impact on the lining plate, so that the grinding efficiency is low, and more importantly, the lining plate can be damaged.

At present, the working state of the domestic ball mill is monitored mainly by workers, and the working states are distinguished by different tones of media balls in different working states. The method not only consumes manpower, but also can not realize all-weather automatic real-time monitoring.

Disclosure of Invention

Based on the problems of the existing monitoring method for the working state of the ball mill, the invention aims to provide a method for detecting the working state of ball mill equipment through monitoring videos, which can not only realize all-weather real-time detection of the working state of the ball mill equipment, but also save labor cost.

The purpose of the invention is realized by the following technical scheme:

the embodiment of the invention provides a method for realizing the detection of the working state of ball mill equipment through a monitoring video, which comprises the following steps:

the outer surface of the cylinder body of the ball mill is provided with a marking unit which is convenient for image recognition processing, and the marking unit can synchronously rotate around the axis of the cylinder body along with the outer surface of the cylinder body;

arranging a camera on the periphery of the ball mill cylinder, and clearly recording a marking operation video of the marking unit rotating synchronously with the ball mill cylinder through the camera;

identifying the motion track of a marking unit in the marking operation video through the image identification processing, converting the identified motion track into a correlation coefficient in real time, and outputting a periodic correlation coefficient curve according to the correlation coefficient;

detecting to obtain the total number of the rising edges of the correlation coefficient curve in unit time, and calculating to obtain the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder during working according to the total number of the rising edges of the correlation coefficient curve in unit time;

and determining the current working state of the detected ball mill according to the calculated size relationship between the actual value and the normal value of any parameter in the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder during working, and performing corresponding processing when the ball mill is in the abnormal working state.

The embodiment of the invention also provides a system for realizing the detection of the working state of the ball mill equipment through the monitoring video, which is used for realizing the method of the invention and comprises the following steps:

the device comprises a marking unit, a camera, an image recognition device, a sampling processing device, a detection processing device and an abnormal state processing device; wherein the content of the first and second substances,

the marking unit is fixedly arranged on the outer surface of the cylinder of the detected ball mill and can synchronously rotate around the axis of the cylinder along with the outer surface of the cylinder;

the camera is arranged on the periphery of the ball mill cylinder and can clearly record a marking operation video of the marking unit rotating synchronously along with the ball mill cylinder;

the image recognition device is respectively in communication connection with the output end of the camera and the sampling processing device, recognizes the motion track of a marking unit in the marking operation video, converts the recognized motion track into a correlation coefficient in real time, and outputs a periodic correlation coefficient curve to the sampling processing device according to the correlation coefficient;

the sampling processing device can detect the correlation coefficient curve, detect the total number of the rising edges of the correlation coefficient curve in unit time, and calculate the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder during working according to the total number of the rising edges of the correlation coefficient curve in unit time;

the detection processing device is in communication connection with the detection processing device and can determine the current working state of the detected ball mill according to the rotating speed, the rotating angular speed and the size relation between the actual value and the normal value of any parameter in the rotating period of the ball mill cylinder during working;

and the abnormal state processing device can perform corresponding processing when the ball mill is in the abnormal working state according to the current working state of the detected ball mill determined by the detection processing device.

According to the technical scheme provided by the invention, the method and the system for detecting the working state of the ball mill equipment through the monitoring video have the beneficial effects that:

the method comprises the steps that a marking unit capable of synchronously rotating around the axis of a cylinder body along with the cylinder body is arranged on the ball mill cylinder body, a marking operation video of the marking unit synchronously rotating around the axis of the cylinder body along with the cylinder body is clearly recorded by a camera arranged outside the ball mill cylinder body, the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder body are determined through identification, sampling and detection of the marking operation video, and the actual working state of the ball mill is further determined; the method realizes that the current working state of the ball mill is detected by accurately measuring the speed of the actual working state of the ball mill under the condition of not influencing production, and abnormal working states such as alarming and manual speed regulation or automatic speed regulation are processed in time. Compared with the mode of manually monitoring the working state of the ball mill, the automatic detection can be realized, and the real-time and accurate detection can be realized.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.

FIG. 1 is a schematic diagram of a low-speed operation state of a ball mill according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a ball mill operating at a normal rotation speed according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of a high-speed operation state of a ball mill according to an embodiment of the present invention;

FIG. 4 is a flowchart of a method for detecting the operating status of a ball mill device through a surveillance video according to an embodiment of the present invention;

fig. 5 is a schematic diagram of a system configuration for implementing the method according to the embodiment of the present invention;

fig. 6 is a schematic diagram of another system configuration for implementing the method according to the embodiment of the present invention;

FIG. 7 is a schematic diagram of a periodic correlation coefficient curve of a ball mill in a critical operating state of normal rotation speed and low rotation speed according to an embodiment of the present invention;

FIG. 8 is a schematic diagram of a periodic correlation coefficient curve of a ball mill in a critical operating state of normal rotation speed and high rotation speed according to an embodiment of the present invention;

the parts corresponding to each mark in the figure are: 1-a labeling unit; 2-a camera; 3-image recognition means; 4-a sampling processing device; 5-detecting and processing the device; 6-abnormal state processing means; 7-ball mill cylinder; 8-bracket.

Detailed Description

The technical solutions in the embodiments of the present invention are clearly and completely described below with reference to the specific contents of the present invention. It is to be understood that the described embodiments are merely exemplary of the invention, and not restrictive of the full scope of the invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention. Details which are not described in detail in the embodiments of the invention belong to the prior art which is known to the person skilled in the art.

As shown in fig. 4, an embodiment of the present invention provides a method for detecting a working state of a ball mill device through a monitoring video, which is a method for detecting a working state of a ball mill quickly and effectively in a speed measurement manner of an image recognition marking unit, so as to alarm and/or regulate a rotation speed of the ball mill, so that the ball mill is in a normal rotation speed working state, and the method includes:

the outer surface of the cylinder body of the ball mill is provided with a marking unit which is convenient for image recognition processing (preferably, the marking unit can adopt a mark with certain size, shape and color which are most beneficial to quick and effective image recognition), and the marking unit can synchronously rotate around the axis of the marking unit along with the outer surface of the cylinder body;

arranging a camera on the periphery of the ball mill cylinder, and clearly recording a marking operation video of the marking unit rotating synchronously with the ball mill cylinder through the camera; preferably, the camera does not need to adopt a special camera, only needs the view field containing the marking unit arranged on the outer surface of the cylinder body, can adopt a monitoring camera for completing other monitoring tasks in a workshop, and only needs the video definition to meet the identification of the marking unit by image identification processing;

identifying the motion track of a marking unit in the marking operation video through the image identification processing, converting the identified motion track into a correlation coefficient in real time, and outputting a periodic correlation coefficient curve according to the correlation coefficient;

detecting to obtain the total number of the rising edges of the correlation coefficient curve in unit time, and calculating to obtain the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder during working according to the total number of the rising edges of the correlation coefficient curve in unit time;

and determining the current working state of the detected ball mill according to the calculated size relationship between the actual value and the normal value of any parameter in the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder during working, and performing corresponding processing when the ball mill is in the abnormal working state.

In the method, the marking unit (see fig. 5 and 6) which is arranged on the outer surface of the ball mill cylinder and facilitates the identification of the image identification process comprises the following steps:

arranging a marking unit on the arc-shaped outer surface of the ball mill cylinder;

or a marking unit is fixedly arranged on the surface of any side of the ball mill cylinder, which deviates from the axis of the ball mill cylinder.

In the method, the periphery of the ball mill cylinder is provided with a camera which comprises the following steps:

and arranging a camera at a position where the marking unit can record the clear video synchronously rotating along with the cylinder on the periphery of the cylinder of the ball mill.

In the method, a marking unit in the marking operation video is identified through the image identification processing, the identified marking unit is converted into a correlation coefficient in real time, and the correlation coefficient is output as a periodic correlation coefficient curve:

presetting matching processing related to the marking unit pixel in the image identification processing, wherein the matching processing can output a correlation coefficient after the marking unit pixel is identified; for example: the method is characterized in that a marking unit arranged on the outer surface of a ball mill cylinder is a red round surface with the diameter of 10 cm, a camera is arranged right opposite to a path through which the marking unit passes, the red round surface is best imaged into a red round spot with the diameter of 50 pixels in a video, a same red round spot with the diameter of 50 pixels is preset in image recognition processing, and a correlation coefficient can be obtained by performing correlation operation on the pixels in a video fixed detection window and the red round spot during image recognition.

According to the correlation coefficient, calculating a correlation coefficient curve through a correlation coefficient curve formula x (t) ═ s (t) + n (t), and outputting the correlation coefficient curve, wherein s (t) is the correlation coefficient of the marking unit pixel in the fixed detection window of the video frame at the time t and the matching processing preset in the image identification processing; n (t) is noise generated by the correlation operation between the non-marked unit pixels in the fixed detection window of the video frame at the time t and the matching process preset in the image identification process, and the n (t) is ignored.

In a frame of video, since the non-marked unit pixels in the fixed detection window are not correlated with the matching module preset in the image recognition device, the noise n (t) is very small and can be ignored. s (t) is relative to the correlation coefficient curve, and along with the rotation of the ball mill cylinder, s (t) is a periodic waveform with a rising edge, and the periodic correlation coefficient curve with the rising edge is output after the image recognition processing. The periodic correlation coefficient curve of the image recognition output is shown in fig. 7 and 8, wherein fig. 7 shows the normal rotation speed and the low rotation speed n ═ n of the ball mill cylinder₁,ω＝ω₁,T_ω＝T₁) The correlation coefficient curve under the critical working state of (1); FIG. 8 shows the normal and high rotation speeds of the ball mill barrel (n ═ n)₂,ω＝ω₂,T_ω＝T₂) The correlation coefficient curve under the critical operating condition of (2).

In the method, the total number of rising edges of the correlation coefficient curve in unit time is detected to be:

at a set sampling period T_s(period T of this_sCan be set as one frame or several frames of time) to sample and output the waveform of the correlation coefficient curve, and obtaining the total number N of the rising edges of the correlation coefficient curve in unit time T from the sampling result;

the rotating speed of the ball mill cylinderWherein T represents a unit time, and N represents the total number of rising edges of the correlation coefficient curve in the unit time T;

the rotation angular velocity of the ball mill cylinderWherein n represents a cylinderThe rotating speed T represents unit time, and N represents the total number of rising edges of the correlation coefficient curve in the unit time T;

the rotation period of the ball mill cylinderWherein N represents the rotation speed, T represents the unit time, and N represents the total number of the rising edges of the correlation coefficient curve in the unit time T.

In the method, the rising edge of the correlation coefficient curve is determined by the following method:

if the difference value of the correlation coefficient curve waveform of the ith sampling point and the last sampling point is calculated_iNot less than a predetermined threshold value₀And then confirming that the sampling point is the rising edge of the correlation coefficient curve, the serial number of the sampling point corresponding to the rising edge of the correlation coefficient curve is i, and accumulating and counting the number of the rising edges of the correlation coefficient curve to obtain the total number N of the rising edges of the correlation coefficient curve in unit time T.

In the method, according to the calculated magnitude relation between the actual value and the normal value of any parameter of the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder body in the working process, the current working state of the detected ball mill is determined, and corresponding processing is carried out when the ball mill is in the abnormal working state at present, wherein the method comprises the following steps:

(1) determining the current working state of the detected ball mill according to the size relation between the actual value and the normal value of the rotating speed of the ball mill cylinder, and performing corresponding processing when the ball mill is in an abnormal working state, wherein the method comprises the following steps:

the rotation speed range of the cylinder body in the normal rotation speed working state of the ball mill to be detected is set as n in advance₁≤n≤n₂Will be described in

The rotating speed range of the cylinder body in the normal rotating speed working state of the detected ball mill is set to be n 1-n 2, and the rotating speed of the ball mill cylinder body is divided into three sets according to different working states:

X_{high speed}＝{n|n＞n₂}

X_{Is normal}＝{n|n₁≤n≤n₂}

X_{Low speed}＝{n|n＜n₁}；

Calculating the actually measured rotating speed n of the ball mill cylinder in the production process, and if the rotating speed n ∈ X is the rotating speed n_{Is normal}Determining that the ball mill is in a normal rotating speed working state, maintaining the current rotating speed, and if the rotating speed is n ∈ X_{High speed}Determining that the ball mill is in a high-rotation-speed working state, and giving out an alarm and/or reducing the rotation speed, if the rotation speed is n ∈ X_{Low speed}Determining that the ball mill is in a low-rotation-speed working state, and performing processing of giving an alarm and/or increasing the rotation speed;

(2) determining the current working state of the detected ball mill according to the size relation between the actual value and the normal value of the rotation angular speed of the ball mill cylinder, and performing corresponding processing when the ball mill is in an abnormal working state, wherein the method comprises the following steps:

the rotation angular speed range of the cylinder body in the normal rotation speed working state of the ball mill to be detected is set to omega in advance₁≤ω≤ω₂Dividing the rotation angular speed of the ball mill cylinder into three sets according to different working states:

Y_{high speed}＝{ω|ω＞ω₂}

Y_{Is normal}＝{ω|ω₁≤ω≤ω₂}

Y_{Low speed}＝{ω|ω＜ω₁}；

Calculating the actually measured rotation angular speed omega of the cylinder body of the ball mill in the production process, and if the rotation angular speed omega ∈ Y_{Is normal}Determining that the ball mill is in a normal rotating speed working state (see figure 2), maintaining the current rotating speed, and if the rotating angular speed omega ∈ Y_{High speed}Determining that the ball mill is in a high-rotation-speed working state (see figure 3), and giving an alarm and/or reducing the rotation speed, if the ball mill rotates at an angular speed omega ∈ Y_{Low speed}Determining that the ball mill is in a low-rotation-speed working state (see figure 1), and performing processing of giving an alarm and/or increasing the rotation speed;

(3) determining the current working state of the detected ball mill according to the size relation between the actual value and the normal value of the rotation period of the ball mill cylinder, and performing corresponding processing when the ball mill is in an abnormal working state, wherein the method comprises the following steps:

the rotation period range of the cylinder body in the normal rotation speed working state of the ball mill to be detected is preset to be T₁≤T_ω≤T₂Dividing the rotation period of the ball mill cylinder into three sets according to different working states:

Z_{high speed}＝{T_ω|T_ω＜T₁}

Z_{Is normal}＝{T_ω|T₁≤T_ω≤T₂}

Z_{Low speed}＝{T_ω|T_ω＞T₂}；

Calculating the actually measured rotation period T of the ball mill cylinder in the production process_ωIf the period of rotation T_ω∈Z_{Is normal}Determining that the ball mill is in a normal rotating speed working state, and maintaining the current rotating speed; if the period of rotation T_ω∈Z_{High speed}Determining that the ball mill is in a high-rotation-speed working state, and performing processing of giving an alarm and/or reducing the rotation speed; if the period of rotation T_ω∈Z_{Low speed}And determining that the ball mill is in a low-rotation-speed working state, and performing processing of giving an alarm and/or increasing the rotation speed.

In the method, the rotating speed range of the cylinder body in the normal rotating speed working state of the detected ball mill is set as n in advance₁≤n≤n₂The rotation speed range of the cylinder body in the normal rotation speed working state of the ball mill to be detected is set as n according to prior knowledge accumulated in the actual production process₁≤n≤n₂；

The rotation angular speed range of the cylinder body in the normal rotation speed working state of the ball mill to be detected is set to omega in advance₁≤ω≤ω₂Setting the rotation angular speed range of the cylinder body in the normal rotation speed working state of the ball mill to be detected as omega according to the prior knowledge accumulated in the actual production process₁≤ω≤ω₂；

The normal rotating speed of the ball mill to be detected is presetThe range of the rotation period of the operating cylinder is set to T₁≤T_ω≤T₂Setting the rotation period range of the cylinder body in the normal rotation speed working state of the ball mill to be detected as T according to the prior knowledge accumulated in the actual production process₁≤T_ω≤T₂。

Referring to fig. 5 and 6, an embodiment of the present invention further provides a system for detecting an operating state of a ball mill device through a surveillance video, so as to implement the method described above, where the system includes:

the device comprises a marking unit, a camera, an image recognition device, a sampling processing device, a detection processing device and an abnormal state processing device; wherein the content of the first and second substances,

the marking unit is fixedly arranged on the outer surface of the cylinder of the detected ball mill and can synchronously rotate around the axis of the cylinder along with the outer surface of the cylinder;

the camera is arranged on the periphery of the ball mill cylinder and can clearly record a marking operation video of the marking unit rotating synchronously along with the ball mill cylinder;

the image recognition device is respectively in communication connection with the output end of the camera and the sampling processing device, recognizes the motion track of a marking unit in the marking operation video, converts the recognized motion track into a correlation coefficient in real time, and outputs a periodic correlation coefficient curve to the sampling processing device according to the correlation coefficient;

the sampling processing device can detect the correlation coefficient curve, detect the total number of the rising edges of the correlation coefficient curve in unit time, and calculate the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder during working according to the total number of the rising edges of the correlation coefficient curve in unit time;

the detection processing device is in communication connection with the detection processing device and can determine the current working state of the detected ball mill according to the rotating speed, the rotating angular speed and the size relation between the actual value and the normal value of any parameter in the rotating period of the ball mill cylinder during working;

and the abnormal state processing device can perform corresponding processing when the ball mill is in the abnormal working state according to the current working state of the detected ball mill determined by the detection processing device.

The image identification device is internally provided with a matching module related to the marking unit, can output a correlation coefficient after identifying the marking unit in the identification mark rotation video, and outputs a periodic correlation coefficient curve according to the correlation coefficient.

In the system, the sampling processing device includes:

the device comprises a sampling device, a counting device and a signal processing device; wherein the content of the first and second substances,

the sampling device is respectively in communication connection with the counting device and the signal processing device and can be used for sampling in a preset sampling period T_sSampling the correlation coefficient curve, and outputting the amplitude of the correlation coefficient curve at each sampling point;

the counting device can calculate the total number of the rising edges of the correlation coefficient curve in unit time T according to the amplitude of the correlation coefficient curve at each sampling point output after sampling by the sampling device;

the signal processing device can obtain the total number of the rising edges of the correlation coefficient curve in the unit time T through the first formula according to the calculation of the counting deviceCalculating the rotating speed N of the ball mill cylinder, wherein in the first formula, T represents unit time, and N represents the total number of rising edges of a correlation coefficient curve in unit time; by a second formulaCalculating to obtain the rotation angular velocity omega of the ball mill, wherein in the second formula, N represents the rotation speed of the ball mill cylinder, T represents unit time, and N represents the total number of rising edges of a correlation coefficient curve in unit time; and by a third formulaIs calculated to obtainThe rotation period T of the ball mill_ωIn the third formula, N represents the rotating speed of the ball mill cylinder, T represents unit time, and N represents the total number of rising edges of the correlation coefficient curve in unit time;

in the system, an alarm circuit and/or a self-control circuit are arranged in the abnormal state processing device, the alarm circuit can give an alarm to the abnormal working state, and the self-control circuit can adjust the rotating speed of the ball mill in the abnormal working state.

Preferably, the camera and the image recognition device of the system can be arranged on a support outside the ball mill cylinder.

The method and the system of the invention estimate and measure the rotating speed of the ball mill cylinder by identifying and converting the monitoring video, further calculate the rotating angular speed and the rotating period of the cylinder, calculate the current working state of the ball mill by utilizing any parameter of the rotating speed, the rotating angular speed and the rotating period, output different signals to the alarm circuit according to the difference of the current working state, control the alarm to give an alarm to workers in the forms of light, sound and the like, remind the workers to change the working rotating speed of the ball mill, and also regulate the speed of the ball mill by the automatic control circuit, so that the ball mill is recovered to the normal working state.

The embodiments of the present invention are described in further detail below.

The embodiment of the invention provides a method for detecting the working state of ball mill equipment by monitoring videos, and as shown in figure 4, the method mainly comprises the following steps:

step S1: the method is characterized in that a marking unit (see fig. 5 and 6) with certain size, shape and color is arranged at a position, convenient for shooting, on the outer surface of the cylinder body of the ball mill, so that the image recognition device can easily perform quick and effective processing, and the marking unit can synchronously rotate around the axis of the marking unit along with the outer surface of the cylinder body;

step S2, arranging a camera at a proper distance outside the ball mill cylinder (the camera does not need a special camera, only needs a view field containing a marking unit arranged on the outer surface of the cylinder, and can simultaneously complete other monitoring tasks of a workshop), and clearly recording a marking rotation video of the marking unit synchronously rotating around the axis of the cylinder along with the cylinder;

step S3, recognizing the marking unit by image recognition processing, converting the recognition result into a correlation coefficient in real time, and outputting a periodic correlation coefficient curve;

the above steps S1 to S3 complete the conversion of the signal form, and convert the video signal reflecting the information of the rotation speed of the ball mill into a correlation coefficient curve convenient for processing, thereby facilitating the subsequent processing. The correlation coefficient curve is: x (t) s (t) + n (t); where s (t) is a correlation coefficient between the pixels in the fixed detection window in the video frame at time t and a matching module (matched with the marking unit) preset in the image recognition apparatus, and n (t) is noise generated by correlation between the pixels in the non-marking unit in the fixed detection window in the video frame at time t and the matching module preset in the image recognition apparatus. Thus, as the ball mill barrel rotates, s (t) is a periodic waveform with a rising edge, and the image recognition device outputs a periodic correlation coefficient curve with a rising edge.

In a frame of video, since the non-marked unit pixels in the fixed detection window are not correlated with the matching module preset in the image recognition device, the noise n (t) is very small and can be ignored. Fig. 7 and 8 show waveforms of the periodic correlation coefficient curve outputted from the image recognition device, wherein fig. 7 shows the normal rotation speed and the low rotation speed n ═ n of the ball mill cylinder₁,ω＝ω₁,T_ω＝T₁) The correlation coefficient curve under the critical working state of (1); FIG. 8 shows the normal and high rotation speeds of the ball mill barrel (n ═ n)₂,ω＝ω₂,T_ω＝T₂) The correlation coefficient curve under the critical operating condition of (2).

Step S4: detecting the total number of the rising edges of the correlation coefficient curve in unit time, and calculating the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder according to the total number of the rising edges of the correlation coefficient curve in unit time;

at a set sampling period T_sSampling and outputting the periodic correlation coefficient curve with rising edge output by the image recognition device (one or several frames of time), and obtaining the unit time T internal phase from the sampling resultThe total number N of rising edges of the correlation coefficient curve;

rotating speed of ball mill barrelWherein T represents unit time, and N represents the total number of rising edges of the relation curve in the unit time T;

angular velocity of rotation of ball mill barrelWherein N represents the rotating speed of the cylinder, T represents unit time, and N represents the total number of the rising edges of the relation number curve in the unit time T;

rotation period of ball mill barrelWhere N represents the rotational speed, T represents the unit time, and N represents the total number of rising edges of the correlation number curve in the unit time T.

In the above method, the total number of the rising edges of the curve representing the relationship number in the unit time T is determined by:

if the difference value of the correlation coefficient curve waveform of the ith sampling point and the last sampling point is calculated_iGreater than a predetermined threshold₀I.e. by_i≥₀And then confirming that the sampling point is the rising edge of the correlation coefficient curve waveform, the serial number of the sampling point corresponding to the rising edge of the correlation coefficient curve waveform is i, and accumulating and counting the number of the rising edges of the correlation coefficient curve waveform to obtain the total number N of the rising edges of the correlation coefficient curve within the unit time T.

In the unit time T of sampling, the rotating speed of the ball mill is considered to be constant,

the rotating speed of the ball mill cylinder is as follows:

the rotating angular speed of the ball mill cylinder is as follows:

the rotation period of the ball mill cylinder is as follows:

step S5: and detecting the working state of the ball mill equipment according to the measured and calculated relationship between the actual value and the normal value of any parameter in the rotating speed, the rotating angular speed and the rotating period of the ball mill cylinder during working in the step S4, and performing corresponding treatment on the abnormal working state, such as alarming the abnormal working state and/or automatically regulating the speed of the ball mill.

Fig. 1 to 3 are schematic diagrams of three operating states of the ball mill, which are determined by the relative relationship between the drop point 30 of the media ball 20 and the material line 10. When the medium ball falling point is slightly below the material line, the medium ball falling point is in a normal rotating speed working state; when the medium ball falling point is far below the material line, the medium ball falling point is in a low-rotation-speed working state; when the medium ball falling point is above the material line, the working state is in a high rotating speed working state. According to physical knowledge, the positions of the throwing point, the dropping point and the material line of the medium ball are all functions of the rotating speed of the ball mill cylinder, and under the condition of the rotating speed of each cylinder, the corresponding relation between the throwing point, the dropping point and the material line of the medium ball is provided.

From the above analysis, three methods are provided for detecting the operating state of the ball mill.

The first method is as follows: according to prior knowledge accumulated in the actual production process, the rotating speed of the cylinder in the normal rotating speed working state of the ball mill is determined within a certain range, and the rotating speed range of the cylinder in the normal working state is set as n₁≤n≤n₂Then, the rotating speed of the ball mill cylinder can be divided into three sets according to different working states:

X_{high speed}＝{n|n＞n₂}

X_{Is normal}＝{n|n₁≤n≤n₂}

X_{Low speed}＝{n|n＜n₁}

Calculating the actually measured rotating speed n of the ball mill cylinder in the rotating speed estimation mode through the monitoring video in the production process, if n ∈ X_{Is normal}The ball mill is in normal rotation speed working state and can maintain the current rotation speed, if n ∈ X_{High speed}If n ∈ X, the ball mill is in high-speed working state and needs to be reduced in speed_{Low speed}If the ball mill is in a low-rotation-speed working state, the rotation speed needs to be increased.

The second method comprises the following steps: according to prior knowledge accumulated in the actual production process, the rotation angular speed of the cylinder in the normal rotation speed working state of the ball mill is determined within a certain range, and the rotation angular speed range of the cylinder in the normal working state is set as omega₁≤ω≤ω₂Then, the rotation angular speed of the ball mill cylinder can be divided into three sets according to different working states:

Y_{high speed}＝{ω|ω＞ω₂}

Y_{Is normal}＝{ω|ω₁≤ω≤ω₂}

Y_{Low speed}＝{ω|ω＜ω₁}

Calculating the actually measured rotation angular speed omega of the ball mill cylinder in the production process by a monitoring video rotation speed estimation mode, if omega ∈ Y_{Is normal}The ball mill is in normal rotation speed working state and can maintain the current rotation speed, if omega ∈ Y_{High speed}If the ball mill is in a high-speed working state, the rotating speed needs to be reduced, and if the rotating speed is omega ∈ Y_{Low speed}If the ball mill is in a low-rotation-speed working state, the rotation speed needs to be increased.

The third method comprises the following steps: according to prior knowledge accumulated in the actual production process, the rotation period of the ball mill in the normal rotation speed working state is determined within a certain range, and the rotation period range of the normal work is set as T₁≤T_ω≤T₂Then, the rotation period can be divided into three sets according to different working states:

Z_{high speed}＝{T_ω|T_ω＜T₁}

Z_{Is normal}＝{T_ω|T₁≤T_ω≤T₂}

Z_{Low speed}＝{T_ω|T_ω＞T₂}

In the process of calculating productionBall mill rotation period T actually measured by monitoring video to estimate rotation speed mode_ωIf T is_ω∈Z_{Is normal}The ball mill is in a normal rotating speed working state, and the current rotating speed is maintained; if T_ω∈Z_{High speed}If the ball mill is in a high-rotation-speed working state, the rotation speed needs to be reduced; if T_ω∈Z_{Low speed}If the ball mill is in a low-rotation-speed working state, the rotation speed needs to be increased.

In conclusion, the rotation speed, the rotation angular speed and the rotation period of the cylinder of the ball mill are obtained in a non-contact mode of monitoring the video to estimate the rotation speed, the current working state of the ball mill can be obtained by utilizing any one of the parameters, different signals are output to the alarm circuit according to the difference of the current working state, the alarm is controlled to give an alarm to workers in the forms of light, sound and the like, the workers are reminded to change the working speed of the ball mill, and further the ball mill can be controlled to regulate the speed according to the current working state through the automatic control circuit.

The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.