CN118111327A - Detection method for preventing secondary loading of main well skip based on laser point cloud deep learning - Google Patents

Abstract

The invention relates to a detection method for preventing secondary loading of a main well skip based on laser point cloud deep learning, which comprises the following steps: establishing a datum point cloud model; waiting for the skip to reach a discharging position signal; acquiring skip position information; screening point cloud data; acquiring point cloud data of a vertical section; calculating the sectional area of the residual material; calculating the volume of the residual materials; a residue evaluation is made. According to the invention, the laser radar arranged outside the skip is utilized to scan the position of the skip, which is most easy to generate the residual material, at the outlet of the skip, so as to obtain point cloud data of the residual material of the skip, analyze the data, obtain the position and the volume of the residual material through calculation, thereby judging the influence of the residual material on the operation of the skip, accurately sending out the residual material quantity and early warning signals in real time, avoiding secondary loading, improving the operation safety of the skip, effectively avoiding overload accidents caused by the secondary loading, and guaranteeing the operation safety of the mine hoist.

Description

Detection method for preventing secondary loading of main well skip based on laser point cloud deep learning

Technical Field

The invention relates to a detection method for preventing secondary loading of a main well skip based on laser point cloud deep learning, which is a detection method in a loading process of transportation equipment and is a loading detection method of a coal mine shaft lifting machine.

Background

The main shaft hoist is used as important large-scale equipment for lifting coal in a coal mine shaft, can meet the requirements of safety, reliability, high efficiency and long-term operation, and can directly influence the normal safety production and economic benefit of the coal mine. Among the factors affecting the safe and reliable operation of the main shaft hoist, skip overload is one of the main factors, and the consequences caused by the skip overload are quite serious. Due to the reasons of limited technical detection means, unfixed running position of equipment and the like, the overload of the main well skip can be reliably detected and radically solved. Therefore, the method has important practical significance in deep research and overload prevention and control technology. In order to ensure the lifting safety of the main well skip, various technical measures for preventing and controlling overload are adopted in each mine, including technical measures such as a fixed weight loading technology, a secondary loading preventing technology, a steel wire rope tension real-time monitoring technology and the like, so that overload and safety accidents caused by overload are avoided to a certain extent. However, the main well has heavy lifting tasks and harsh physical environment, and the raw coal produced by the coal mine is influenced by moisture and gangue content, so that the change of the volume weight and viscosity is relatively large, and the phenomenon of skip unloading coal stagnation, namely 'coal not unloaded', is unavoidable, thereby generating overload and even 'secondary loading'. The term "secondary loading" means: when the skip is lifted to the wellhead for unloading, the coal in the skip is not completely unloaded, and a certain amount of coal remains in the skip to close the unloading gate, so that the skip is unloaded again after being lifted to the bottom of the well.

The prior method for detecting the 'coal not unloaded' mainly comprises the following steps of detecting by a 'post detection' method, namely lifting and leaving an unloading position after receiving a skip, and obtaining a conclusion of the 'coal not unloaded' by comparing motor current changes or wire rope tension changes generated by the weight difference of the skip when the skip is in a constant-speed operation stage, and sending out early warning before the next loading. The detection method mainly has two problems, namely, the false alarm condition caused by the change of the loading capacity exists, and the detection means belongs to 'post detection', namely, when the fact that the coal is not unloaded is found, an unloaded skip (heavy skip) is in a descending operation state, and the production efficiency is influenced when the heavy load is lowered and potential safety hazards exist. The other method for detecting overload in real time is to add a tension sensor at the position where the steel wire rope is connected with the skip, and the method can acquire the weight of the skip in real time so as to judge whether the skip is unloaded or overloaded, but the method has high cost, and because the tension sensor reciprocates up and down along with the skip, the power supply and data transmission are difficult, the failure rate and maintenance amount are high, and the method is difficult to effectively popularize and apply. Therefore, finding a reliable and economical method for detecting the phenomenon of "coal not being discharged" is a problem to be solved.

Disclosure of Invention

In order to overcome the problems in the prior art, the invention provides a detection method for preventing secondary loading of a main well skip based on laser point cloud deep learning. According to the method, after the skip is unloaded, the laser radar is utilized to scan and detect the coal storage condition at the bottom of the skip, and the coal storage quantity and the early warning signal are accurately sent in real time so as to stop secondary loading.

The purpose of the invention is realized in the following way: a detection method for preventing secondary loading of a main well skip based on laser point cloud deep learning, wherein a system used by the method comprises the following steps: at least one laser radar arranged outside a skip discharging outlet when the skip is stopped at a discharging position, wherein the laser radar is electrically connected with a point cloud data calculation server and a skip lifting controller in sequence; the method comprises the following steps:

Step 1, building a datum point cloud model: the method comprises the steps of obtaining point cloud data of a discharge opening of an emptying skip, learning reference data, and establishing a reference point cloud model of the discharge opening, wherein the process is as follows:

1) Scanning a skip discharge opening which is positioned at a discharge position and is completely discharged and the lower part of the skip which is positioned below the discharge opening by using a laser radar to obtain laser point cloud data;

2) Establishing a three-dimensional coordinate system, and taking a laser radar as a space 0 point;

3) Searching a space coordinate position of the lower edge of the skip from the point cloud data, and a plurality of key points in an unloading port of the skip;

4) Establishing a discharge port reference data model by key points;

Step 2, waiting for skip to reach a discharging position signal: after the skip lifting controller detects that the skip is in place and discharges the skip, a skip in-place signal is sent to the point cloud data calculation server;

Step 3, acquiring skip position information: after receiving the skip in-place signal, the point cloud data calculation server starts point cloud data screening, and the position of the lower edge of the skip is obtained through a skip lower edge identification method;

step 4, screening point cloud data: screening the point cloud data of the inner space of the current discharge opening through machine learning;

Step 5, acquiring point cloud data of a vertical section: subdividing the section inside the discharge opening according to the laser beam by using a vertical section, and screening laser point clouds with a plurality of sections;

step 6, calculating the sectional area of the residual materials: performing polygonal area calculation on each section point cloud;

Step 7, calculating the volume of the residual materials: carrying out calculus summation on the sectional areas to calculate the volume of the residual materials;

Step 8, making a residue evaluation: and (5) evaluating the volume of the residual materials and sending a conclusion to the skip lifting controller.

The invention has the advantages and beneficial effects that: according to the invention, the laser radar arranged outside the skip is utilized to scan the position of the skip, which is most easy to generate the residual material, at the outlet of the skip, so as to obtain point cloud data of the residual material of the skip, analyze the data, obtain the position and the volume of the residual material through calculation, thereby judging the influence of the residual material on the operation of the skip, accurately sending out the residual material quantity and early warning signals in real time, avoiding secondary loading, improving the operation safety of the skip, effectively avoiding overload accidents caused by the secondary loading, and guaranteeing the operation safety of the mine hoist.

Drawings

The invention is further described below with reference to the drawings and examples.

FIG. 1 is a schematic diagram of a system and a skip thereof used in the method according to the embodiment of the present invention, and a coordinate of 0 point is a laser radar, which is a cross-sectional view of a skip elevation;

FIG. 2 is a flow chart of a method according to an embodiment of the invention;

FIG. 3 is a front view of a laser radar scanning skip discharge port in accordance with an embodiment of the present invention, and is a view of the arrow B in FIG. 1;

FIG. 4 is a side view coordinate system of a laser radar scanning skip of the method of an embodiment of the present invention, which is a schematic elevation view of the skip of FIG. 1;

FIG. 5 is a graph of the point cloud profile of the skip discharge outlet in xy coordinate system for a residual curve of approximately L2, z=0 for a discharge cycle of the method of the present invention;

FIG. 6 is a schematic diagram of coordinates of a cross-sectional area of a sweep under xy coordinates at a skip discharge port for a discharge cycle of the method of an embodiment of the present invention;

figure 7 is a schematic representation of the differential area of the cross-section of the residue at a certain discharge cycle of the method according to an embodiment of the invention.

Detailed Description

Examples:

the embodiment is a detection method for preventing secondary loading of a main well skip based on laser point cloud deep learning, wherein a system used by the method comprises the following steps: at least one laser radar 2 arranged outside the skip discharge outlet 101 when the skip 1 is stopped at the discharge position, wherein the laser radar is arranged at a position right below the discharge opening, and the laser radar can scan the inside of the discharge opening in the maximum area. The laser radar is electrically connected with the point cloud data calculation server and the skip lifting controller in sequence, as shown in figure 1; the connection may be by a wired or wireless network, such as ethernet, etc.

The skip of the present embodiment is a container for carrying coal or other ore for lifting mined coal or ore from the bottom of the well to the surface. The skip according to this embodiment is a high cylinder with a rectangular horizontal section and a discharge hole at one side of the bottom, the bottom is provided with an inclined plate 102, materials enter the skip from the upper part of the high cylinder, and are discharged from the discharge hole at one side of the bottom of the skip under the action of the inclined plate, as shown by arrow A in FIG. 1, to indicate the flow process of the coal. Of course, other forms of skip can be used, and the method described in the embodiment can be used for detecting as long as the laser radar arranged outside the skip discharge hole can irradiate the residual materials inside the skip.

And (3) placing 1 or 2 multi-line laser radars at a certain position right opposite to the discharge opening at the bottom of the skip, scanning the discharge opening below the skip by using the laser radars, and acquiring three-dimensional point cloud data of the discharge opening and the inside of the discharge opening. Comparing and matching the three-dimensional data of the laser point cloud of the discharge opening after discharging with the point cloud data of the discharge opening which is normally discharged to obtain a conclusion of the bonding condition of the discharge opening below the skip, and comprising the following steps: plugging, partial sticking, unloading, etc.

The point cloud data calculation server and the skip lifting controller can be general industrial computers or other electronic devices with digital operation and storage capacity, such as a PLC (programmable logic controller).

One difficulty with measuring skips is: because the unloading position is not fixed (when overhauling every day, overhauling workers can adjust according to production working conditions) and the steel wire rope stretches and changes along with the working time, the unloading position of the skip bucket is caused to have a certain change range (+ -0.4 m), and the point cloud is difficult to match (the unloading port is positioned). The method comprises the steps of positioning the lower edge of a skip by using a laser radar, measuring the deviation of the skip from a standard unloading position (data in the skip running direction), translating datum point cloud data (coal unloading point cloud) according to the deviation data, and then comparing and matching.

The principle of scanning skip residual materials and calculating residual material quantity by using a laser radar is shown in fig. 1, and the area where the laser radar can radiate and scan through a discharge opening is shown as a shaded part when seen from a side sectional view of the skip. As long as the coal is detected in the area, the amount of the residual coal can be estimated according to the distribution data of the coal measurement points. The stock curves L1, L2, L3 represent three possible intersecting lines of the stock surface and the skip section, and the part overlapping with the shadow is the position of the landing point of the laser point on the residual material, such as the small circle mark part in fig. 1. The cross-sectional area of the residual scanned by the laser scanning line is not difficult to obtain by curve fitting.

The specific steps of the method are as follows, and the flow is shown in fig. 2:

Step 1, building a datum point cloud model: the method comprises the steps of obtaining point cloud data of a discharge opening of an emptying skip, learning reference data, and establishing a reference point cloud model of the discharge opening, wherein the process is as follows:

1) And scanning a skip discharge opening which is positioned at a discharge position and is completely discharged and the lower part of the skip which is positioned below the discharge opening by using a laser radar to obtain laser point cloud data.

2) A three-dimensional rectangular coordinate system is established, and a laser radar is taken as an origin of the coordinate system, see fig. 3 and 4.

3) Searching a space coordinate position of the lower edge of the skip from the point cloud data, and a plurality of key points in an unloading port of the skip; six discharge port reference data key points are shown in P0, P2, P3, P5, P7 and P8 in FIG. 3 and FIG. 4.

And slicing and calculating the volume of the residual materials according to laser lines falling on the residual materials by utilizing a laser radar multi-line scanning mechanism. To measure the spatial volume, a spatial three-dimensional coordinate system is first established.

A three-dimensional coordinate system is established by taking the space position of the laser radar as a coordinate origin 0, fig. 3 is a zy coordinate system (the paper surface is a zy plane) of a front view skip of the laser radar, fig. 4 is an xy coordinate system (the paper surface is an xy plane), a longitudinal section of a certain discharge opening of the skip is selected, and z=0 is set. Different z values can be determined from different laser lines, the more laser beams scanned to the discharge opening, the more slices can be divided. Under the zy coordinate system, laser points P7 and P8 of the empty skip in a static state are obtained in advance, wherein the two points are the intersection points of the extension lines of two vertical edges of the discharge opening and the lower edge of the skip along the line, and two numerical values Z ₇ and Z ₈ are obtained. Because the left edge and the right edge of the skip cannot deviate when the skip is unloaded at the wellhead, the Z value range of the laser point falling on the coal in the unloading opening is Z ₇<z_i<Z₈, and the range is one of the laser point cloud screening conditions.

4) Establishing a discharge port reference data model by key points; the model is used for matching and comparing the screening conditions of the point cloud in the discharge opening, such as a formula (1), and the analysis is as follows:

Under the xy coordinate system when z=0, laser points P0, P1, P2, P3 and P4 of the empty skip in a static state are obtained in advance, wherein P0 is a skip lower edge point, P1 is a skip coal chute plate lower edge, P2 is a skip coal chute plate upper edge, P3 is a skip coal discharge opening upper edge, and P4 is a certain point above the skip discharge opening. And the coordinates of P1 and P3 can be used for obtaining a space coordinate curve of the coal chute plate, and the P3 point can be used for obtaining P5, namely, a point where P3 falls on the oblique coal chute plate along the y-axis direction. It is thus easy to derive that the laser spot falling on the coal in the discharge opening has an X value range of X ₃<x_i<X₂ and a y value range of The range is two laser point cloud screening conditions.

Step 2, waiting for a skip to reach a discharging position (and finishing discharging) signal: and after the skip lifting controller detects that the skip is in place and discharges, sending a skip in place signal to the point cloud data calculation server.

Step 3, acquiring skip position information: after receiving the skip in-place signal, the point cloud data calculation server starts point cloud data screening, and the skip lower edge position p0 is obtained through a skip lower edge identification method.

The datum point cloud data of the discharge opening comprise P0, P3, P5 and P2, the diagonal line section P5P2, and the screening condition of the discharge opening point cloud (datum point cloud) is as follows:

And after the datum point cloud of the discharge opening is provided, carrying out offset matching on the discharge opening of the real-time skip. Because the skip discharge position is not always fixed.

Step 4, screening point cloud data: screening the point cloud data of the inner space of the current discharge opening through machine learning; the front discharge opening inner space point cloud data is residue point cloud, and the screening method is as shown in formula (2).

Point cloud data acquisition (xy coordinate system of a certain laser beam when z=0) is performed on the skip discharge port of a certain discharge cycle, as shown in fig. 5. Firstly, finding a point P0 (x ₀,y₀) of the lower edge of the skip by an algorithm, wherein the offset of the point P0 of the lower edge of the skip and the reference point P0 of the discharge opening is (y ₀-Y₀), and the cloud data P0', P3', P5', P2' of the reference point of the discharge opening under the discharge cycle and the inclined line segment P5'P2' are easily deduced. Thereby can obtain the screening condition of this circulation lower discharge opening inside point cloud:

The screening conditions of equation 2 can find the laser point p1 (x ₁,y₁),p2(x₂,y₂)...pN(x_N,y_N) inside the discharge opening from the laser beam laser point cloud, and calculate the area of the coal under the section from these laser points, as shown in the ghost part of fig. 5. For easy solution, the calculation can be performed as the ghost part shown in fig. 6, that is, the pN point is projected to the skip back plate in parallel to the oblique line P2' P5', so that the Y coordinate of the point P6' is obtained, which is (Y ₂+y₀-Y₀+y_N-y₁). The volume of the residue can be obtained by integrating the polygon P1 (x 1, y 1), P2 (x 2, y 2). PN (xN, yN), P2', P6' calculus to obtain the area S ₁, and then integrating the values of a plurality of tangential planes of the discharge port point cloud in the z-axis direction, i.e., v=Σs _i ×Δz.

Step 5, acquiring point cloud data of a vertical section: and carrying out section subdivision on the inside of the discharge opening in a z-axis coordinate according to a laser beam by using a vertical section, and screening laser point clouds with a plurality of sections.

Step 6, calculating the sectional area of the residual materials: performing polygons P1 (x 1, y 1), P2 (x 2, y 2) & pN (xN, yN), P2', P6' to find the area Si for each cross-sectional point cloud; the area is calculated by the micro-integration method, and the algorithm process is shown in fig. 6 and 7.

Step 7, calculating the volume of the residual materials: and carrying out calculus summation on the sectional areas to calculate the volume of the residual materials.

For the calculus area S ₁, P1 (x 1, y 1), P2 (x 2, y 2)..pn (xN, yN) can be projected to the skip back plate in parallel to the diagonal line P2'P5', and the area S _i can be calculated respectively for a plurality of polygons differentiated, finally S ₁＝Σs_i is obtained. As shown in fig. 7, according to the two-dimensional coordinate system polygon area solving method (the solving area of the plurality of trapezoids), it is not difficult to obtain a subdivision polygon area S _i having a certain cross-sectional area:

the algorithm process for finding the lower edge point of the skip is as follows: the minimum value Y _min of the lower edge position of the skip after unloading is manually determined, and the range Y _min<y_i<(Y_min +0.8 of the lower edge of the skip unloading position can be determined according to the fact that the upper and lower floating range of the skip unloading position is +/-0.4 meter. For example, the y value of the position where the lower edge of the skip is located after unloading is at least-0.2 m, and then the y value range of the lower edge of the skip can be determined to be-0.2 to-0.6 m according to the fact that the up-down floating range of the unloading position of the skip is +/-0.4 m. The laser point cloud is screened according to the condition of formula 4, and the laser point p0 can be found.

It should be noted that: the discharge position of each cycle of skip discharge is changed only in the y-axis direction, and the x-axis and the z-axis are not offset.

Step 8, making a residue evaluation: and (5) evaluating the volume of the residual materials and sending a conclusion to the skip lifting controller.

And evaluating the size and the position of the residual materials according to the calculation result to judge the damage degree of the residual materials, judging whether the residual materials are required to be cleaned, and sending the evaluation result to the skip controller so that the skip controller can perform actions such as continuous operation, alarm or stop operation.

Finally, it should be noted that the above is only for illustrating the technical solution of the present invention and not for limiting the same, and although the present invention has been described in detail with reference to the preferred arrangement, it should be understood by those skilled in the art that the technical solution of the present invention (such as the form of the skip, the application of various formulas, the sequence of steps, etc.) may be modified or replaced equivalently without departing from the spirit and scope of the technical solution of the present invention.

Claims (1)

1. A detection method for preventing secondary loading of a main well skip based on laser point cloud deep learning, wherein a system used by the method comprises the following steps: at least one laser radar arranged outside a skip discharging outlet when the skip is stopped at a discharging position, wherein the laser radar is electrically connected with a point cloud data calculation server and a skip lifting controller in sequence; the method is characterized by comprising the following steps:

Step 1, building a datum point cloud model: the method comprises the steps of obtaining point cloud data of a discharge opening of an emptying skip, learning reference data, and establishing a reference point cloud model of the discharge opening, wherein the process is as follows:

1) Scanning a skip discharge opening which is positioned at a discharge position and is completely discharged and the lower part of the skip which is positioned below the discharge opening by using a laser radar to obtain laser point cloud data;

2) Establishing a three-dimensional coordinate system, and taking a laser radar as a space 0 point;

3) Searching a space coordinate position of the lower edge of the skip from the point cloud data, and a plurality of key points in an unloading port of the skip;

4) Establishing a discharge port reference data model by key points;

Step 2, waiting for skip to reach a discharging position signal: after the skip lifting controller detects that the skip is in place and discharges the skip, a skip in-place signal is sent to the point cloud data calculation server;

Step 3, acquiring skip position information: after receiving the skip in-place signal, the point cloud data calculation server starts point cloud data screening, and the position of the lower edge of the skip is obtained through a skip lower edge identification method;

step 4, screening point cloud data: screening the point cloud data of the inner space of the current discharge opening through machine learning;

Step 5, acquiring point cloud data of a vertical section: subdividing the section inside the discharge opening according to the laser beam by using a vertical section, and screening laser point clouds with a plurality of sections;

step 6, calculating the sectional area of the residual materials: performing polygonal area calculation on each section point cloud;

Step 7, calculating the volume of the residual materials: carrying out calculus summation on the sectional areas to calculate the volume of the residual materials;

Step 8, making a residue evaluation: and (5) evaluating the volume of the residual materials and sending a conclusion to the skip lifting controller.