A crane lifting point calculation method and system based on scrap steel shape recognition

Abstract

The embodiment of the application discloses a kind of based on scrap steel shape identification's overhead travelling crane hoisting point calculation method and system, comprising: collecting the original point cloud data of scrap steel shape in scrap steel pool;Read in the basic automation data and hoisting plan information of overhead travelling crane;Original point cloud data and basic automation data are input into scrap steel shape identification model and are handled, and output scrap steel surface shape standardization data;Scrap steel surface shape standardization data are input into hoisting point calculation software, and output overhead travelling crane hoisting point, start hoisting operation.The embodiment of the application, original point cloud data and basic automation data are input into scrap steel shape identification model, and carry out grid division, neighborhood search, neighborhood data mean and interpolation processing, output scrap steel surface shape standardization data, realize the accurate positioning of scrap steel;Through hoisting point calculation software, realize the hoisting point calculation based on scrap steel shape identification, and then start hoisting operation, realize through scrap steel logistics intelligentization, unmanned power enterprise development.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of scrap steel unmanned crane, and particularly relates to a crane hoisting point calculation method and system based on scrap steel shape identification. BACKGROUND

[0002] The steel enterprises have frequent cross operations of the scrap steel in and out of the warehouse, which requires the cooperation of crane workers and warehouse management personnel.

[0003] In the process of the scrap steel in and out of the warehouse, the identification of the shape of the scrap steel and the selection of the crane hoisting point still need manual participation, and such traditional operation form is low in efficiency, high in risk and high in labor cost.

[0004] Therefore, it is a big problem to be solved in the field that the unmanned crane system automatically identifies the shape of the scrap steel and automatically calculates the crane hoisting point to reduce manual participation and reduce risks. SUMMARY

[0005] In order to solve the above technical problems, the present application provides the following technical solutions:

[0006] In the first aspect, the present application provides a crane hoisting point calculation method based on scrap steel shape identification, comprising:

[0007] Collecting original point cloud data of the shape of scrap steel in a scrap steel pool;

[0008] Reading in basic automation data and hoisting plan information of the crane;

[0009] Inputting the original point cloud data and the basic automation data into a scrap steel shape identification model for processing to output scrap steel surface shape standardized data;

[0010] Inputting the scrap steel surface shape standardized data and the hoisting plan information into hoisting point calculation software to output a crane hoisting point;

[0011] Starting a hoisting operation according to the crane hoisting point.

[0012] Further, the original point cloud data of the shape of scrap steel in the scrap steel pool is collected by a laser scanner rotating scanning mode.

[0013] Further, the inputting of the original point cloud data and the basic automation data into the scrap steel shape identification model for processing to output the scrap steel surface shape standardized data comprises:

[0014] Generating first data by preprocessing the original point cloud data;

[0015] Generating second data by coordinate transformation of the first data according to the basic automation data;

[0016] griding the second data to generate cell data;

[0017] neighbor searching the second data to generate neighborhood data;

[0018] averaging the neighborhood data to generate averaged data;

[0019] interpolating missing data in the averaged data to generate normalized data of the surface shape of the scrap steel.

[0020] Further, the griding the second data to generate cell data comprises:

[0021] establishing a minimum bounding box of all the second data;

[0022] calculating a step length for dividing the minimum bounding box according to a number of the second data in a preset single cell, a total number of the second data, and the minimum bounding box;

[0023] dividing the cell data of the minimum bounding box in a three-dimensional space according to the minimum bounding box and the step length.

[0024] Further, the neighbor searching the second data to generate neighborhood data comprises:

[0025] calculating a cell serial number of any of the second data in the minimum bounding box according to the cell data;

[0026] finding K points closest to any of the second data from the cell serial number and a first preset number of neighborhood cells, and marking the K points as the neighborhood data;

[0027] Otherwise,

[0028] if the first preset number of neighborhood cells does not find the K points closest to any of the second data, expanding to a second preset number of neighborhood cells to find the K points closest to any of the second data, and marking the K points as the neighborhood data.

[0029] Further, the inputting the normalized data of the surface shape of the scrap steel and the hoisting plan information into the hoisting point calculation software to output the crane hoisting points comprises:

[0030] inputting the normalized data of the surface shape of the scrap steel into the hoisting point calculation software;

[0031] selecting a hoisting point calculation mode,

[0032] if the hoisting point calculation mode is an accurate mode, outputting a third preset number of crane hoisting points according to the accurate mode,

[0033] Otherwise,

[0034] If the lifting point calculation mode is the fast mode, output the third preset number of crane lifting points according to the fast mode.

[0035] Further, output the third preset number of crane lifting points according to the fast mode, comprising:

[0036] Select the highest point of the scrap steel;

[0037] According to the highest point of the scrap steel, select the corresponding scrap steel surface shape standardized data within the range of the crane suction cup, and calculate the lowest point of the scrap steel;

[0038] According to the elevation, the highest point and the lowest point of the scrap steel, calculate the difference and tolerance of the scrap steel elevation;

[0039] According to the difference and tolerance of the scrap steel elevation, output the third preset number of crane lifting points.

[0040] Further, output the third preset number of crane lifting points according to the fast mode, comprising:

[0041] Select the highest point of the scrap steel;

[0042] According to the highest point of the scrap steel, select the fourth preset number of scrap steel surface shape standardized data within the range of the crane suction cup corresponding to the highest point, and calculate the difference of the scrap steel elevation;

[0043] According to the difference of the scrap steel elevation, output the third preset number of crane lifting points.

[0044] In a second aspect, the embodiments of the present application provide a crane lifting point calculation system based on scrap steel shape identification, comprising:

[0045] A data acquisition module is configured to acquire original point cloud data of the shape of the scrap steel in the scrap steel pool;

[0046] A basic automation module is configured to read in basic automation data and lifting plan information of the crane;

[0047] A point cloud data processing module is configured to input the original point cloud data and the basic automation data into a scrap steel shape identification model for processing, and output scrap steel surface shape standardized data;

[0048] A lifting point calculation module is configured to input the scrap steel surface shape standardized data and the lifting plan information into a lifting point calculation software, and output crane lifting points;

[0049] An automatic lifting control module is configured to start a lifting operation according to the crane lifting points.

[0050] Further, the point cloud data processing module comprises:

[0051] a preprocessing submodule, configured to preprocess the original point cloud data to generate first data;

[0052] a coordinate conversion submodule, configured to perform coordinate conversion on the first data according to the basic automation data to generate second data;

[0053] a grid division submodule, configured to perform grid division on the second data to generate cell data;

[0054] a neighborhood search calculation submodule, configured to perform neighborhood search on the second data to generate neighborhood data, and perform mean value processing on the neighborhood data to generate mean value data;

[0055] an interpolation processing submodule, configured to perform interpolation processing on missing data in the mean value data to generate normalized data of the surface shape of the scrap steel.

[0056] Further, the system further comprises a crown travel positioning module, which monitors the position of the crown travel and other crown positions in the system through a laser displacement sensor on a crown travel mechanism.

[0057] Beneficial effects:

[0058] (1) The method and system for calculating the hoisting point of the crown based on the shape identification of the scrap steel according to the embodiment of the application input the original point cloud data and the basic automation data into the scrap steel shape identification model, perform grid division, neighborhood search, neighborhood data mean value processing and interpolation processing, output normalized data of the surface shape of the scrap steel, and then realize accurate positioning of the scrap steel; input the normalized data of the surface shape of the scrap steel and the hoisting plan information into hoisting point calculation software to realize hoisting point calculation based on the shape identification of the scrap steel; start the hoisting operation according to the hoisting point of the crown to realize intelligent and unmanned scrap steel logistics to help enterprise development.

[0059] (2) The method and system for calculating the hoisting point of the crown based on the shape identification of the scrap steel according to the embodiment of the application comprehensively utilize technologies such as laser scanning imaging and ranging to quickly scan the entire profile of the scrap steel pool area to form three-dimensional data; quickly distinguish the scanning area during the scanning process to realize identification of the shape of the scrap steel through a three-dimensional modeling method. The method and system according to the embodiment of the application realize unmanned operation and automatic running of the crown, thereby reducing the demand for management personnel for the operation of the crown and the ground storage area, improving labor efficiency, and ensuring personal safety, thereby reducing costs and increasing benefits for enterprises. BRIEF DESCRIPTION OF DRAWINGS

[0060] Figure 1 The method flowchart provided by the embodiment of the application;

[0061] Figure 2A raw point cloud data calculation schematic diagram under a laser scanner coordinate system provided for an embodiment of the present application is shown in FIG. 1.

[0062] Figure 3 A flowchart of inputting raw point cloud data and basic automation data into a scrap steel shape recognition model for processing and outputting scrap steel surface shape standardized data is shown in FIG. 2.

[0063] Figure 4 A flowchart of grid division on second data and generation of single cell data is shown in FIG. 3.

[0064] Figure 5 A system structure block diagram provided for an embodiment of the present application is shown in FIG. 4.

[0065] Figure 6 A point cloud data processing module structure block diagram provided for an embodiment of the present application is shown in FIG. 5. DETAILED DESCRIPTION

[0066] The present application will be described below in conjunction with the accompanying drawings and specific embodiments.

[0067] Figure 1 A method flowchart provided for an embodiment of the present application is shown in FIG. 6, which includes:

[0068] Raw point cloud data of scrap steel shape in a scrap steel pool is collected;

[0069] Basic automation data and hoisting plan information of a crown block are read in;

[0070] Raw point cloud data and basic automation data are input into a scrap steel shape recognition model for processing, and scrap steel surface shape standardized data is output;

[0071] Scrap steel surface shape standardized data and hoisting plan information are input into hoisting point calculation software, and a crown block hoisting point is output;

[0072] According to the crown block hoisting point, a hoisting operation is started.

[0073] The realization of scrap steel three-dimensional shape recognition is based on laser scanning and crown block operation data, and raw point cloud data of scrap steel shape in a scrap steel pool is collected through laser scanner rotation scanning. When a hoisting operation is started, a laser scanner scans the scrap steel pool once, and after recognizing the scrap steel shape through a three-dimensional modeling method, a set number of hoisting points are calculated. After the hoisting points are hoisted, scanning is performed again until the hoisting is completed.

[0074] A scrap steel recognition module communicates with a laser scanner through Ethernet to obtain laser scanning data. The scrap steel recognition module communicates with a basic automation module through OPC to read crown block data, and simultaneously sends the crown block data to a scrap steel shape recognition model at a regular time for processing, and outputs scrap steel surface shape standardized data.

[0075] The laser scanner is a diffuse reflection type, and the shape profile of the scrap steel in the scrap steel pool is measured by using a laser ranging principle. The original point cloud data acquired by the laser scanner in a single frame is in a linear distribution, so a motion mechanism is needed to pull and move regularly to realize complete measurement of the surface of the scrap steel. Considering the field situation, the embodiment of the application adopts a rotating mode, uses a servo motor system and a laser scanner integrated scanning device, adopts a crane beam installation mode, and installs a two-dimensional laser scanner on the mechanism. The laser scanner main body is installed on the rotatable integrated scanning device to measure the shape and position information of the scrap steel.

[0076] Before the overhead crane lifting operation, the overhead crane is moved to the vicinity of the center position of the scrap steel pool, and the integrated scanning device is started. After the laser scanner reads the walking position of the trolley, the integrated scanning device is started. During the operation of the integrated scanning device, the laser scanner moves uniformly from left to right along the Y-axis direction, and the encoder can feed back the angle in real time. During the uniform rotation of the laser scanner from left to right, the laser scanner simultaneously scans in the front-to-back direction at a frequency of 190 degrees and 35 Hz. After the scanning of the laser scanner along the Y-axis direction is completed, the integrated scanning device returns to the original position.

[0077] The laser scanner scans the scrap steel to acquire the position information of each point on the surface of the scrap steel, i.e. the original point cloud data, and transmits the original point cloud data to the data acquisition module and the point cloud data processing module of the industrial computer.

[0078] In the embodiment of the application, the laser scanning imaging, ranging and other technologies are comprehensively utilized to realize unmanned operation and automatic running of the overhead crane, reduce the demand for management personnel of the overhead crane operation and the ground warehouse area, improve the labor efficiency, and ensure the personal safety, thereby reducing the cost and increasing the efficiency for the enterprise.

[0079] Figure 2 The original point cloud data calculation schematic diagram under the coordinate system of the laser scanner provided by the embodiment of the application is provided, the laser scanner is set as the coordinate origin O(0, 0, 0), and the original point cloud data coordinates (X mk , Y mk , Z mk ) of the scrap steel are calculated by the formula:

[0080] X mk =R mk *cosβ*cos(θ0+k*θ r ) (1)

[0081] Y mk =R mk *sinβ (2)

[0082] Z mk =R mk *cosβ*sin(θ0+k*θr ) (3)

[0083] wherein R mk is the straight-line distance of the scrap steel from the coordinate origin, β is the pan-tilt swing angle of the scrap steel relative to the coordinate origin, θ0 is the starting angle of the laser scanner, k is the scan point serial number, and θ r is the angular resolution.

[0084] To establish a three-dimensional model of the scrap steel, the current running position of the unmanned trolley is obtained, and the original point cloud data in the laser scanner coordinate system is converted into coordinate values in the scrap steel span coordinate system. Since the refresh frequency of the laser scanner and the refresh frequency of the encoder data are inconsistent, the refresh frequency of the encoder data (generally 1024 frames / s) is greater than the refresh frequency of the laser scanner, so time matching of the data is required. The matching principle is based on the time of the laser scanning data, and the closest encoder data is used to match the laser scanning data.

[0085] The basic automation module reads in basic automation data, including the current running value of the trolley, the position of the trolley, etc. The output data of the basic automation module is the coordinate value of the trolley hoisting point.

[0086] Figure 3 The flowchart provided by the embodiment of the present application for inputting the original point cloud data and the basic automation data into the scrap steel shape recognition model, processing, and outputting the standardized data of the scrap steel surface shape, includes:

[0087] The original point cloud data is preprocessed to generate first data;

[0088] The first data is coordinate-converted according to the basic automation data to generate second data;

[0089] The second data is grid-divided to generate cell data;

[0090] The second data is neighborhood-searched to generate neighborhood data;

[0091] The neighborhood data is mean-processed to generate mean data;

[0092] The missing data in the mean data is interpolated to generate standardized data of the scrap steel surface shape.

[0093] After the original point cloud data is preprocessed to filter out interference data caused by shaking and shielding to generate first data, the coordinate conversion is performed in combination with the size data of the trolley and the installation position data of the laser scanner, so that the coordinates of the measurement points are converted from the laser scanner coordinate system to the scrap steel span coordinate system.

[0094] Since the original point cloud data of laser scanning is relatively dense, in order to reduce the amount of rendering calculation when constructing a three-dimensional model, the second data is simplified by using the grid division method to generate cell data.

[0095] The original point cloud data is a collection of scattered points, and the relationship between the data cannot be determined. Therefore, after the second data is gridded, the neighborhood relationship between the points is established.

[0096] Figure 4 The flowchart of gridding the second data and generating cell data is provided in an embodiment of the present invention. The embodiment of the present invention adopts the cell method as the space division algorithm. The principle is to evenly divide the point cloud data in each direction using a step length, and the resulting small spaces are all cubes with a side length equal to the step length. The steps are as follows:

[0097] ① Establish the minimum bounding box of all second data,

[0098] Search all the second data and find the point P with the smallest coordinates min (x min 、y min 、z min ) and the point P with the largest coordinates max (x max 、y max 、z max ), determine a cube containing all point cloud data based on these two points, and this cube is the minimum bounding box;

[0099] ② According to the number of second data in a preset single cell, the total number of second data and the minimum bounding box, calculate the step size of dividing the minimum bounding box,

[0100] Let the step size be d, and divide the minimum bounding box into n small cells according to this step size:

[0101]

[0102] In formula (4), Indicates rounding up.

[0103] The number of second data in a single cell is preset to k, and the total number N of second data is expressed as:

[0104] N=akn (5)

[0105] According to formulas (4) and (5), the step length d is:

[0106]

[0107] Wherein, k is generally 6 or 26 according to experience, represents the neighborhood of the cell, a is an adjustable parameter, and a suitable step is determined by adjusting a;

[0108] III. According to the minimum bounding box and the step, the cell data of the minimum bounding box in the three-dimensional space is divided into,

[0109]

[0110] In formula (7), L x , L y , L z are the cell data calculated on each coordinate axis.

[0111] According to the neighborhood search on the second data, neighborhood data is generated, including:

[0112] According to the cell data, the cell serial number of any second data in the minimum bounding box is calculated;

[0113] From the cell serial number and the first preset number of neighborhood cells, the K nearest points to the any second data are searched, and are sorted according to the distance and marked as neighborhood data;

[0114] Otherwise,

[0115] If the K nearest points to the any second data are not found in the first preset number of neighborhood cells, the K nearest points to the any second data are searched in the second preset number of neighborhood cells, and are marked as neighborhood data.

[0116] According to the cell data, the cell serial number of any second data P i (P ix , P iy , P iz in the minimum bounding box is calculated as,

[0117]

[0118] After filtering out the interference data, a size-varying hole is generated in the point cloud data, and when a three-dimensional surface model is constructed, the neighborhood data needs to be processed by mean value to generate mean value data, the missing data in the mean value data is processed by interpolation to generate normalized data (digital elevation model data) of the scrap steel surface shape, and is stored in a database.

[0119] The method of the embodiment of the application forms a plurality of grids with equal size and uniform distribution in the effective stacking range of the scrap steel span after grid division, realizes digital segmentation of the scrap steel span, and realizes seamless image splicing by normalizing and integrating the number of points of the measuring points falling in the grids.

[0120] The normalized data of the surface shape of the scrap steel and the hoisting plan information are input into the hoisting point calculation software, and the crane hoisting points are output, including:

[0121] The normalized data of the surface shape of the scrap steel and the hoisting plan information are input into the hoisting point calculation software;

[0122] The hoisting point calculation mode is selected,

[0123] If the hoisting point calculation mode is the accurate mode, the third preset number of crane hoisting points are output according to the accurate mode,

[0124] Otherwise,

[0125] The hoisting point calculation mode is the fast mode, and the third preset number of crane hoisting points are output according to the fast mode.

[0126] The third preset number of crane hoisting points are output according to the accurate mode, including:

[0127] The highest point of the scrap steel is selected;

[0128] According to the highest point of the scrap steel, the normalized data of the surface shape of the scrap steel in the corresponding crane suction disc range is selected, and the lowest point of the scrap steel is calculated;

[0129] According to the elevation of the scrap steel, the highest point and the lowest point, the difference value and the tolerance of the scrap steel elevation are calculated;

[0130] According to the difference value and the tolerance of the scrap steel elevation, the third preset number of crane hoisting points are output.

[0131] In the accurate mode, if the difference value and the tolerance of the scrap steel elevation meet the preset value, the highest point of the scrap steel is marked as the first hoisting point, and the other normalized data of the surface shape of the scrap steel in the corresponding crane suction disc range is marked as the selected hoisting point. After filtering out the first hoisting point and the selected hoisting point, the highest point of the scrap steel is selected again, and the selection and marking of the second hoisting point are performed again. Otherwise, if the difference value and the tolerance of the scrap steel elevation do not meet the preset value, the highest point of the scrap steel is marked as an unhoistable point, and after filtering out the unhoistable point, the highest point of the scrap steel is selected again, and the selection and marking of the first hoisting point are performed again.

[0132] According to the third preset number of crane hoisting point selection requirements, the steps of selecting the highest point of the scrap steel, judging whether the difference value and the tolerance of the scrap steel elevation meet the preset value, and selecting and marking the next hoisting point are repeated until the number of hoisting points reaches the requirement, and the selection and calculation are ended. When the number of calculations exceeds the third preset number, an alarm is given to remind that the number of hoisting points selected in the accurate mode cannot reach the set number.

[0133] After the selection of the hoisting points is completed, the hoisting process is started after being confirmed by the operator in the automatic hoisting control module, a hoisting task is generated, and the hoisting task is automatically issued to the PLC of the corresponding overhead crane.

[0134] According to the quick mode, a third preset number of hoisting points of the overhead crane are output, including:

[0135] The highest point of the scrap steel is selected.

[0136] According to the highest point of the scrap steel, a fourth preset number of scrap steel surface shape normalization data in the corresponding range of the suction cup of the overhead crane are selected, and a difference in the elevation of the scrap steel is calculated.

[0137] According to the difference in the elevation of the scrap steel, a third preset number of hoisting points of the overhead crane are output.

[0138] The fourth preset number can be four corners of the range of the suction cup of the overhead crane, or other numbers of corners of the range of the suction cup of the overhead crane.

[0139] In the quick mode, if the difference in the elevation of the scrap steel meets a preset value, the highest point of the scrap steel is marked as a first hoisting point, and other scrap steel surface shape normalization data in the range of the suction cup of the overhead crane corresponding to the highest point of the scrap steel are marked as selected hoisting points; after the first hoisting point and the selected hoisting points are filtered out, the highest point of the scrap steel is selected again, and the selection and marking of the second hoisting point are performed again; otherwise, if the difference in the elevation of the scrap steel does not meet the preset value, the highest point of the scrap steel is marked as an unhoistable point, and after the unhoistable point is filtered out, the highest point of the scrap steel is selected again, and the selection and marking of the first hoisting point are performed again.

[0140] According to the selection requirement of the third preset number of hoisting points of the overhead crane, the steps of selecting the highest point of the scrap steel, judging whether the difference in the elevation of the scrap steel meets the preset value, and selecting and marking the next hoisting point are repeated until the number of the hoisting points reaches the requirement, and the selection and calculation are ended. When the number of calculations exceeds the third preset number, an alarm is given to remind that the selection of the hoisting points cannot reach the set number in the quick mode.

[0141] After the selection of the hoisting points is completed, the hoisting process is started after being confirmed by the operator in the automatic hoisting control module, a hoisting task is generated, and the hoisting task is automatically issued to the PLC of the corresponding overhead crane.

[0142] In the method of the embodiment of the present application, the surface shape normalized data of the scrap steel and the hoisting plan information are input into a hoisting point calculation software to generate a crane hoisting point. After receiving the hoisting plan information (the hoisting plan information can be created through an interface or received from a steelmaking production management system plan) containing the scanning system number, the scrap steel pool number, the work type and the like, the basic automation module looks up the current use state of the corresponding scrap steel pool in the basic information database according to the hoisting plan information, and according to the selection of the work machine and the hoisting point calculation mode by the operator, the crane hoisting point and the hoisting control parameter are output in the hoisting point calculation module according to the surface shape normalized data of the scrap steel.

[0143] Figure 5 The system structure block diagram provided by the embodiment of the present application comprises:

[0144] The data acquisition module is used for acquiring the original point cloud data of the scrap steel shape in the scrap steel pool.

[0145] The basic automation module is used for reading in the basic automation data of the crane and the hoisting plan information.

[0146] The point cloud data processing module is used for inputting the original point cloud data and the basic automation data into a scrap steel shape recognition model for processing to output the surface shape normalized data of the scrap steel.

[0147] The hoisting point calculation module is used for inputting the surface shape normalized data of the scrap steel and the hoisting plan information into a hoisting point calculation software to output the crane hoisting point.

[0148] The automatic hoisting control module is used for starting the hoisting operation according to the crane hoisting point.

[0149] The basic automation module establishes a basic automation database containing a plurality of basic information such as scrap steel spans and scrap steels to provide support for generating an automatic hoisting model.

[0150] Figure 6 The point cloud data processing module structure block diagram provided by the embodiment of the present application comprises:

[0151] The preprocessing submodule is used for preprocessing the original point cloud data to generate first data.

[0152] The coordinate conversion submodule is used for performing coordinate conversion on the first data according to the basic automation data to generate second data.

[0153] The grid division submodule is used for performing grid division on the second data to generate cell data.

[0154] The neighborhood search calculation submodule is used for performing neighborhood search on the second data to generate neighborhood data, performing mean value processing on the neighborhood data to generate mean value data.

[0155] An interpolation processing submodule is configured to interpolate missing data in the mean data to generate normalized data of the surface shape of the scrap steel.

[0156] In the system, a man-machine interface, a hoisting operation interface and a scrap steel span maintenance interface are arranged.

[0157] The hoisting operation interface is used for viewing the three-dimensional shape of the scrap steel span, creating and modifying a device operation plan, issuing an operation task and the like.

[0158] The system further comprises a crown block walking positioning module.

[0159] In the system, a laser displacement sensor is used to position the walking of the crown block, so that the cumulative effect caused by errors of a traditional absolute value encoder is avoided, and relative accuracy of positioning is achieved.

[0160] The method and system can realize intelligent automatic control of the crown block on the basis of fully exerting the hoisting capacity and other functions of the crown block.

[0161] It is to be noted that, in the present text, relational terms such as "first" and "second", and the like, can only be used to distinguish one entity or action from another entity or action, without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises a... " does not, without more constraints, exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.

Claims

1. A crane hoisting point calculation method based on scrap steel shape recognition, characterized in that, The method comprises: Collecting original point cloud data of the shape of scrap steel in a scrap steel pool; Reading in basic automation data and hoisting plan information of a crane; Inputting the original point cloud data and the basic automation data into a scrap steel shape recognition model for processing, and outputting normalized data of the surface shape of the scrap steel; Inputting the normalized data of the surface shape of the scrap steel and the hoisting plan information into hoisting point calculation software, and outputting crane hoisting points, comprising: Inputting the normalized data of the surface shape of the scrap steel and the hoisting plan information into hoisting point calculation software; Selecting a hoisting point calculation mode, If the hoisting point calculation mode is an accurate mode, outputting a third preset number of crane hoisting points according to the accurate mode, otherwise, If the hoisting point calculation mode is a fast mode, outputting a third preset number of crane hoisting points according to the fast mode; outputting a third preset number of crane hoisting points according to the accurate mode, comprising: Selecting a highest point of the scrap steel; According to the highest point of the scrap steel, selecting the normalized data of the surface shape of the scrap steel within the corresponding crane suction cup range, and calculating the lowest point of the scrap steel; According to the elevation of the scrap steel, the highest point and the lowest point, calculating the difference and tolerance of the elevation of the scrap steel; According to the difference and tolerance of the elevation of the scrap steel, outputting a third preset number of crane hoisting points; According to the fast mode, outputting a third preset number of crane hoisting points, comprising: Selecting a highest point of the scrap steel; According to the highest point of the scrap steel, selecting a fourth preset number of the normalized data of the surface shape of the scrap steel within the corresponding crane suction cup range, and calculating the difference of the elevation of the scrap steel; According to the difference of the elevation of the scrap steel, outputting a third preset number of crane hoisting points; According to the crane hoisting points, starting the hoisting operation.

2. The crane hoisting point calculation method based on scrap steel shape recognition according to claim 1, characterized in that, Inputting the original point cloud data and the basic automation data into a scrap steel shape recognition model for processing, and outputting normalized data of the surface shape of the scrap steel, comprising: Pretreating the original point cloud data to generate first data; According to the basic automation data, performing coordinate transformation on the first data to generate second data; Grid dividing the second data to generate cell data; Neighborhood searching of the second data to generate neighborhood data; Mean value processing of the neighborhood data to generate mean value data; Interpolation processing of missing data in the mean value data to generate normalized data of the surface shape of the scrap steel.

3. The method for shape recognition based scrap steel to determine the crane lifting point according to claim 2, characterized in that, Grid dividing the second data to generate cell data, comprising: Establishing a minimum bounding box of all the second data; According to the number of second data in a preset single cell, the total number of second data and the minimum bounding box, calculating the step length for dividing the minimum bounding box; According to the minimum bounding box and the step length, dividing the cell data of the minimum bounding box in the three-dimensional space.

4. The method for shape recognition based scrap steel to determine the crane lifting point according to claim 3, characterized in that, The neighborhood searching of the second data to generate neighborhood data, comprising: According to the cell data, calculating the cell serial number of any second data in the minimum bounding box; from the cell serial number and a first preset number of neighborhood cells, finding the K nearest points to any second data, and marking as neighborhood data; Otherwise, If the K nearest points to any of the second data are not found in the first preset number of neighborhood cells, the search is expanded to the second preset number of neighborhood cells to find the K nearest points to any of the second data, which are marked as neighborhood data.

5. A crown block hoisting point calculation system based on scrap steel shape recognition, characterized in that, The system comprises: A data acquisition module for acquiring original point cloud data of the shape of the scrap steel in the scrap steel pool; A basic automation module for reading in basic automation data and hoisting plan information of the overhead crane; A point cloud data processing module for inputting the original point cloud data and the basic automation data into a scrap steel shape recognition model for processing, and outputting normalized data of the surface shape of the scrap steel; A hoisting point calculation module for inputting the normalized data of the surface shape of the scrap steel and the hoisting plan information into a hoisting point calculation software, and outputting hoisting points of the overhead crane, comprising: inputting the normalized data of the surface shape of the scrap steel and the hoisting plan information into the hoisting point calculation software; selecting a hoisting point calculation mode, if the hoisting point calculation mode is the accurate mode, outputting a third preset number of hoisting points of the overhead crane according to the accurate mode, otherwise, if the hoisting point calculation mode is the fast mode, outputting a third preset number of hoisting points of the overhead crane according to the fast mode; outputting a third preset number of hoisting points of the overhead crane according to the accurate mode, comprising: selecting a highest point of the scrap steel; selecting the normalized data of the surface shape of the scrap steel within the corresponding range of the suction disc of the overhead crane according to the highest point of the scrap steel, and calculating a lowest point of the scrap steel; calculating a difference value and a tolerance of the elevation of the scrap steel according to the elevation, the highest point and the lowest point of the scrap steel; outputting a third preset number of hoisting points of the overhead crane according to the difference value and the tolerance of the elevation of the scrap steel; outputting a third preset number of hoisting points of the overhead crane according to the fast mode, comprising: selecting a highest point of the scrap steel; selecting a fourth preset number of the normalized data of the surface shape of the scrap steel within the corresponding range of the suction disc of the overhead crane according to the highest point of the scrap steel, and calculating a difference value of the elevation of the scrap steel; outputting a third preset number of hoisting points of the overhead crane according to the difference value of the elevation of the scrap steel; An automatic hoisting control module for starting a hoisting operation according to the hoisting points of the overhead crane.

6. The scrap shape recognition based crown block hoisting point calculation system according to claim 5, wherein, The point cloud data processing module comprises: a preprocessing submodule for preprocessing the original point cloud data to generate first data; a coordinate conversion submodule for converting the coordinates of the first data according to the basic automation data to generate second data; a grid division submodule for dividing the second data into cells to generate cell data; a neighborhood search calculation submodule for searching the neighborhood of the second data to generate neighborhood data, and performing mean value processing on the neighborhood data to generate mean value data; an interpolation processing submodule for interpolating missing data in the mean value data to generate normalized data of the surface shape of the scrap steel.

7. The shape recognition based scrap steel based crane hoisting point calculation system according to claim 5, wherein, The system further comprises an overhead crane travel positioning module for monitoring the travel position of the overhead crane and the positions of other overhead cranes in the system through a laser displacement sensor on a travel mechanism of the overhead crane.

Images