A laser vision-based rod wire rolling force measuring method and measuring device

Abstract

The present application relates to a kind of laser vision-based bar wire rolling force measurement method and device, the method includes the following steps: using different rolling force, simulate the bar wire of different shaft diameter is rolled;And in simulation process, collect the oil pressure and rolling force data of multiple groups of oil film bearing for supporting rolling mill;Using visual identification method, collect the shaft diameter size data of multiple groups of bar wire before and after rolling;Oil pressure and shaft diameter size data are used as input, and rolling force data are used as output, establish the parameter mapping neural network model based on CNN;Multiple groups of oil pressure, shaft diameter size and rolling force data are used as training set, train the parameter mapping neural network model based on CNN, obtain the network model for predicting rolling force;In actual rolling process, the oil pressure of oil film bearing and the shaft diameter size before and after bar wire rolling are monitored, and the network model for predicting rolling force is used to predict rolling force.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of rolling technology, in particular to a rod wire rolling force measurement method and device based on laser vision. BACKGROUND

[0002] The steel industry refers to the industry of producing pig iron, steel, steel products, industrial pure iron and iron alloy, and is one of the basic industries in all industrialized countries in the world. In 2021, China's crude steel output exceeded 1 billion tons, of which the output of rod wire accounted for 40%-50% on average, and was widely used in bearings, springs, welding rods, mechanical equipment, building materials and other aspects. The rod wire rolling equipment is a rolling mill specially used for rolling rod wire. Due to the small size of the wire section, the total elongation coefficient is large during the rolling process from the blank to the finished product, and a large amount of hot rolling process equipment is needed. Generally, it is composed of 21 to 28 continuous rolling mills, with the highest production speed reaching 140 m / s, and the rod wire temperature close to 900℃. During the rolling process, the vertical size of the rolling force will affect the finished product quality of the rod wire. If the rolling force is too small, it will lead to poor surface quality and uneven deformation of the finished product. If the rolling force is too large, it will increase the rolling waste, additional energy and material consumption. Therefore, the rolling force monitoring during the rolling process is of great significance to ensure product quality and process adjustment.

[0003] The existing plate and strip rolling equipment can directly monitor the rolling force by installing a special measurement system, such as the rolling force measurement system of ABB and KELK company, which directly measures the rolling force by arranging pressure sensors between the bearing seat, rack and nut. However, most of the existing rod wire rolling equipment is limited by the roll box structure and high temperature and high speed rolling conditions, and it is difficult to directly measure the rolling force by arranging sensors. More and more researchers are focusing on predicting the rolling force through model construction to achieve indirect measurement. The traditional rolling force prediction model calculates the rolling force by combining the key parameters of the relevant sensors through material theory. However, due to the difficulty in obtaining some parameters, there is a deviation between the calculation result and the actual value. The method of using a neural network model to predict the rolling force needs to obtain the actual rolling force during network training, which makes it difficult to build the model.

[0004] For example, Chinese patent CN111400928B discloses a rolling force compensation method and device based on multiple regression. The method is based on a rolling force calculation model and historical rolling data, and iteratively calculates the rolling force compensation coefficient through a multiple linear regression model to realize the compensation of the rolling force.

[0005] A rolling force compensation method and system suitable for a cold continuous rolling mill are disclosed in Chinese Patent CN112588840B. The method compensates the tension of the cold continuous rolling mill during the speed-up and speed-down stages by calculating the additional tension based on the equipment parameters and rolling process parameters of the cold continuous rolling mill. Such methods are generally applicable to plate and strip rolling processes, and the model accuracy needs to be verified. The measuring device is difficult to adapt to high-speed, high-temperature harsh environments, making it difficult to achieve accurate measurement of rod and wire rolling forces. SUMMARY

[0006] The technical problem to be solved by the present application is to provide a rod and wire rolling force measurement method and device based on laser vision to address the above shortcomings. The present application collects relevant data through a simulation test device and constructs a model related to rolling force based on the relevant data, thereby achieving indirect measurement of rod and wire rolling force.

[0007] To solve the above technical problems, the present application adopts the following technical solutions:

[0008] A rod and wire rolling force measurement method based on laser vision, comprising the following steps:

[0009] Different rolling forces are used to simulate the rolling of rod and wire with different shaft diameters;

[0010] During the simulation process, the oil pressure of the oil film bearing supporting the rolling mill and the rolling force data are collected, and the shaft diameter size data before and after rolling of the rod and wire are collected using a visual recognition method;

[0011] The oil pressure and shaft diameter size data are used as input, and the rolling force data are used as output to establish a CNN-based parameter mapping neural network model;

[0012] The multiple sets of oil pressure, shaft diameter size and rolling force data are used as a training set to train the CNN-based parameter mapping neural network model to obtain a network model for predicting rolling force;

[0013] During actual rolling, the oil pressure of the oil film bearing and the shaft diameter size before and after rolling of the rod and wire are monitored, and the rolling force is predicted by the network model for predicting rolling force.

[0014] Further, the process of collecting the shaft diameter size before and after rolling of the rod and wire by visual recognition method includes calibration, image correction and light profile recognition.

[0015] Further, the calibration and image correction include the following steps:

[0016] The black and white grid intersection points on the calibration plate are used as calibration reference points;

[0017] The geometric change matrix is used to correspond one-to-one the calibration reference points on the calibration board and the calibration reference points in the image captured by the camera;

[0018] The geometric change matrix of the camera calibration is obtained by solving the geometric change matrix using the corresponding calibration reference points.

[0019] Further, the light line profile recognition includes the following steps:

[0020] The laser is used to irradiate the calibration board, and the rod wire before and after rolling is located between the laser and the calibration board to form the profile projection of the rod wire before and after rolling on the calibration board;

[0021] The camera is used to capture the calibration board, the profile projection image of the rod wire before and after rolling is collected, and the Hough change is used to calculate the profile size of the rod wire before and after rolling in the profile projection image;

[0022] The profile size is mapped through the geometric change matrix of the camera calibration to obtain the actual profile size, so as to obtain the shaft diameter size of the rod wire before and after rolling.

[0023] Further, the CNN-based parameter mapping neural network model includes three ResNet structures, and the ResNet structure is composed of two convolution layers and two pooling layers alternately combined.

[0024] A rod wire rolling force measuring device based on laser vision includes a rolling simulation unit, a visual recognition unit and a data acquisition unit,

[0025] The rolling simulation unit includes a roller box, a gear box and a pressing mechanism, the roller box includes a first roller shaft and a second roller shaft, the gear box circumscribes a power unit, the gear box is used to transmit power to the first roller shaft, and the pressing mechanism is used to apply pressure to the second roller shaft towards the first roller shaft;

[0026] The visual recognition unit is used to collect the profile image of the rod wire before and after rolling, and the shaft diameter size of the rod wire before and after rolling is obtained according to the profile image;

[0027] The data acquisition unit is used to collect the oil pressure of the oil film bearing on the first roller shaft, and the data acquisition unit is also used to collect the pressure applied by the pressing mechanism to the second roller shaft.

[0028] Further, the visual recognition unit includes a laser, a camera and a calibration board, the laser is used to emit laser to the calibration board, the rod wire to be rolled is located between the laser and the calibration board, and the camera is used to collect the profile image on the calibration board.

[0029] Further, the laser propagation direction of the laser is perpendicular to the calibration board;

[0030] The camera obliquely photographs the calibration board.

[0031] Further, the data acquisition device comprises an oil pressure sensor and a pressure sensor, the oil pressure sensor is used for detecting the oil pressure of the oil film bearing, and the pressure sensor is used for monitoring the pressure output by the pressurizing mechanism.

[0032] Compared with the prior art, the above technical scheme has the following advantages:

[0033] (1) The present application collects the shaft diameter before and after the rod wire rolling through visual recognition, and this kind of collection method is not affected by high speed and high temperature working conditions;

[0034] (2) After obtaining the related data of the rolling process, the present application realizes the acquisition of rolling force through a neural network model, which can be used for quality monitoring and process adjustment.

[0035] The present application will be described in detail below in combination with the drawings and examples. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 It is a general layout diagram of the measuring device;

[0037] Figure 2 It is a partial structure diagram of the measuring device;

[0038] Figure 3 It is a data acquisition schematic diagram of the measuring device;

[0039] Figure 4 It is a structure schematic diagram of the visual recognition unit;

[0040] Figure 5 It is a principle diagram of the visual recognition unit;

[0041] Figure 6 It is a structure schematic diagram of the parameter mapping neural network model based on CNN.

[0042] In the drawings, the components represented by each reference numeral are listed as follows:

[0043] 1, roll box; 11, first roller shaft; 12, second roller shaft; 2, gear box; 3, pressurizing mechanism; 4, oil film bearing; 51, laser; 52, camera; 53, calibration board; 61, oil pressure sensor; 62, pressure sensor; 7, power unit; 8, rod wire. DETAILED DESCRIPTION

[0044] The principles and characteristics of the present application are described below in combination with the drawings, and the examples are only used to explain the present application, and are not used to limit the scope of the present application.

[0045] In the description of the present application, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", "clockwise" and "counterclockwise" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application.

[0046] Embodiment one:

[0047] A laser vision-based rod wire rolling force measurement method, comprising the following steps:

[0048] S1: Building a laser vision-based rod wire rolling force measurement device, which includes a rolling simulation unit, a visual recognition unit and a data acquisition unit;

[0049] S2: Using the laser vision-based rod wire rolling force measurement device to simulate the rolling process of the rod wire;

[0050] S3: During the simulation process, the average oil pressure P 1m and the average oil pressure P 2m of the upper and lower oil film bearings of the first roller shaft 11 are collected by the data acquisition unit, and the rolling force F output by the pressing mechanism 3 is collected by the visual recognition unit, and the shaft diameter size d1 before rolling and the shaft diameter size d2 after rolling are collected by the visual recognition unit;

[0051] S4, repeat steps S2-S3 several times, and change the rolling force F output by the pressing mechanism 3 and the shaft diameter of the rod wire to be rolled, to obtain multiple groups of P 1m , P 2m , F, d1 and d2 data;

[0052] S5, taking P 1m , P 2m , d1 and d2 as input, and F as output, a CNN-based parameter mapping neural network model is established;

[0053] S6, using the multiple groups of P 1m , P 2m , F, d1 and d2 data obtained in step S4 as a training set, the CNN-based parameter mapping neural network model is trained to obtain a network model for predicting the rolling force;

[0054] S7, in the actual rolling process, the input / output shaft diameter of the rod wire and the oil pressure data of the upper / lower oil film bearing are monitored in real time, and the rolling force is predicted by the network model for predicting the rolling force.

[0055] AsFigure 6 As shown, the structure diagram of the parameter mapping neural network model of the CNN, wherein, conv is a convolution operation, max_pooling is a pooling operation, concat is a tensor splicing operation, batch_normalization is a batch normalization operation, dense is a tensor direct connection operation, and flatten is a two-dimensional tensor to one-dimensional array expansion operation.

[0056] In the embodiment, the parameter mapping neural network model of the CNN mainly consists of three ResNet structures, and the three ResNet structures use the idea of hierarchical extraction. Each ResNet structure is responsible for information extraction at the corresponding level. The ResNet structure is formed by alternately combining two convolution layers and two pooling layers, and is mainly used for hierarchical abstraction of data in the current ResNet structure.

[0057] The processing flow of the parameter mapping neural network model of the CNN includes the following steps:

[0058] (a1) combining the input data into a set of two-dimensional data and inputting;

[0059] (a2) abstracting the data in the current ResNet structure by the ResNet structure in a hierarchical manner, and splicing the top layer data and the output data of the current ResNet structure;

[0060] (a3) repeating step (2) three times, and performing a batch normalization operation between each ResNet structure to normalize and connect the front and rear ResNet structures;

[0061] (a4) after completing three times of feature extraction of the ResNet structure, performing two times of batch normalization and tensor connection, and then performing two-dimensional tensor expansion processing and outputting the result.

[0062] The process of collecting the shaft diameter size d1 before rolling and the shaft diameter size d2 after rolling of the rod and wire by the visual recognition unit includes calibration, image correction and light contour recognition, and specifically includes the following steps:

[0063] (b1) taking the black and white grid intersection points on the calibration board as the calibration reference points;

[0064] (b2) using a geometric transformation matrix to one-to-one correspond the calibration reference points on the calibration board with the calibration reference points in the camera image;

[0065] (b3) using the multiple sets of corresponding calibration reference points in step (b2) to solve the geometric transformation matrix to obtain the geometric transformation matrix of the camera calibration;

[0066] (b4) using a laser to irradiate the calibration plate, and the rod wire before and after rolling is located between the laser and the calibration plate, so as to form the profile projection of the rod wire before and after rolling on the calibration plate;

[0067] (b5) using a camera to shoot the calibration plate, collecting the profile projection image of the rod wire before and after rolling, and calculating the profile size of the rod wire before and after rolling in the profile projection image by using Hough change;

[0068] (b6) mapping the profile size by the geometric change matrix of the camera calibration to obtain the actual profile size, so as to obtain the shaft diameter size of the rod wire before and after rolling.

[0069] Regarding the average oil pressure P of the oil film bearing 4 1m and the average oil pressure P of the lower oil film bearing 2m The acquisition of the average oil pressure of the oil film bearing, when setting the data acquisition unit, a plurality of oil pressure sensors 61 are arranged uniformly on the oil film bearing, the plurality of oil pressure sensors will measure a plurality of oil pressure data, and the plurality of oil pressure data is averaged, and the average value obtained is the average oil pressure of the corresponding oil film bearing.

[0070] Example two:

[0071] Described in this embodiment is a measuring device for implementing the method in example one.

[0072] A rod wire rolling force measuring device based on laser vision, comprising a rolling simulation unit, a visual recognition unit and a data acquisition unit, wherein:

[0073] The rolling simulation unit comprises a roller box 1, a gear box 2 and a pressing mechanism 3, the roller box 1 comprises a first roller shaft 11 and a second roller shaft 12, the gear box 2 is circumscribed by a power unit 7, the power unit 7 is a motor, the gear box 2 is used to transmit power to the roller box 1, so as to drive the first roller shaft 11 to rotate, the pressing mechanism 3 is used to apply pressure (i.e. rolling force) to the second roller shaft 12 towards the first roller shaft 11, the pressure mechanism 3 can be a hydraulic pressing mechanism, and the rod wire 8 to be rolled is arranged between the first roller shaft 11 and the second roller shaft 12.

[0074] The visual recognition unit comprises a laser 51, a camera 52 and a calibration plate 53, the calibration plate 53 is arranged on the roller box 1, the calibration plate 53 is provided with a black and white checkerboard image, the propagation direction of the laser emitted by the laser 51 is perpendicular to the surface of the calibration plate 53, and the rod wire 8 to be rolled is located between the laser 51 and the calibration plate, so that the profile of the rod wire 8 is projected onto the calibration plate 53, and the camera 52 is used to shoot the calibration plate 53 obliquely at a certain angle, so as to collect the profile projection image of the rod wire 8;

[0075] The first roller shaft 11 and the second roller shaft 12 are each provided with a set of visual identification units, that is, a set of visual identification units are arranged at the input and output ends of the bar wire, respectively, for collecting the projection images before and after the bar wire rolling.

[0076] The data acquisition unit includes an oil pressure sensor 61 and a pressure sensor 62, the upper part and the lower part of the first roller shaft 11 are each provided with a set of oil film bearings 4 and ball bearings, a plurality of oil pressure sensors 61 are uniformly arranged on the oil film bearings 4, and the average value of all the oil pressure sensors 61 is taken after the data of the oil pressure sensors 61 is collected; the pressure sensor 62 is arranged at the rear end of the pressing mechanism 3, and the pressure sensor 62 is used to measure the pressure (i.e. rolling force) output by the pressing mechanism 3.

[0077] The data acquisition unit includes an oil pressure sensor 61 and a pressure sensor 62, the upper part and the lower part of the first roller shaft 11 are each provided with a set of oil film bearings 4 and ball bearings, a plurality of oil pressure sensors 61 are uniformly arranged on the oil film bearings 4, and the average value of all the oil pressure sensors 61 is taken after the data of the oil pressure sensors 61 is collected; the pressure sensor 62 is arranged at the rear end of the pressing mechanism 3, and the pressure sensor 62 is used to measure the pressure (i.e. rolling force) output by the pressing mechanism 3.

[0078] The above is an example of the best embodiment of the present application, wherein the parts not described in detail are all the common knowledge of ordinary skilled in the art. The protection scope of the present application is subject to the content of the claims, and any equivalent transformation based on the technical inspiration of the present application is also within the protection scope of the present application.

Claims

1. A method for measuring the rolling force of bar and wire rods based on laser vision, characterized in that, Includes the following steps: Different rolling forces were used to simulate rolling bars and wires with different shaft diameters; During the simulation, multiple sets of oil pressure and rolling force data of the oil film bearings used to support the rolling rolls were collected, and multiple sets of shaft diameter data of bar and wire rods before and after rolling were collected using visual recognition methods. Using hydraulic pressure and shaft diameter data as inputs and rolling force data as outputs, a parameter mapping neural network model based on CNN is established. Multiple sets of hydraulic pressure, shaft diameter, and rolling force data were used as training sets to train a CNN-based parameter mapping neural network model, resulting in a network model for predicting rolling force. In the actual rolling process, the oil pressure of the oil film bearing and the shaft diameter of the bar and wire rod before and after rolling are monitored, and the rolling force is predicted by a network model used to predict the rolling force.

2. The method for measuring the rolling force of bar and wire rods based on laser vision according to claim 1, characterized in that, The process of acquiring the shaft diameter of bar and wire rods before and after rolling using visual recognition methods includes calibration, image correction, and light profile recognition.

3. The method for measuring the rolling force of bar and wire rods based on laser vision according to claim 2, characterized in that, The calibration and image correction include the following steps: The intersection of the black and white squares on the calibration plate is used as the calibration reference point; Using a geometric transformation matrix, a one-to-one correspondence is established between the calibration reference points on the calibration plate and the calibration reference points in the images captured by the camera; The geometric transformation matrix is ​​obtained by solving the geometric transformation matrix using the above-mentioned multiple sets of corresponding calibration reference points.

4. The method for measuring the rolling force of bar and wire rods based on laser vision according to claim 3, characterized in that, The light contour recognition includes the following steps: A laser is used to irradiate a calibration plate, with the bar and wire before and after rolling located between the laser and the calibration plate, so as to form the profile projection of the bar and wire before and after rolling on the calibration plate. A camera is used to photograph the calibration plate, and contour projection images of the bar and wire rod before and after rolling are acquired. The contour dimensions of the bar and wire rod before and after rolling are calculated in the contour projection images using Hough variation. The profile dimensions are mapped by the geometric transformation matrix of camera calibration to obtain the actual profile dimensions, thereby obtaining the shaft diameter dimensions of the bar and wire rod before and after rolling.

5. The method for measuring the rolling force of bar and wire rods based on laser vision according to claim 1, characterized in that, The CNN-based parameter mapping neural network model includes three ResNet structures, each consisting of an alternating combination of two convolutional layers and two pooling layers.

6. A laser vision-based bar and wire rolling force measuring device, used to implement the laser vision-based bar and wire rolling force measuring method as described in claim 1, characterized in that, It includes a rolling simulation unit, a vision recognition unit, and a data acquisition unit. The rolling simulation unit includes a roll box (1), a gearbox (2) and a pressurizing mechanism (3). The roll box (1) includes a first roll shaft (11) and a second roll shaft (12). The gearbox (2) is connected to an external power unit (7). The gearbox (2) is used to transmit power to the first roll shaft (11). The pressurizing mechanism (3) is used to apply pressure to the second roll shaft (12) toward the first roll shaft (11). The visual recognition unit is used to acquire contour images of the bar and wire before and after rolling, and to obtain the shaft diameter of the bar and wire before and after rolling based on the contour images. The data acquisition unit is used to acquire the oil pressure of the oil film bearing (4) on the first roller shaft (11), and the data acquisition unit is also used to acquire the pressure applied by the pressurizing mechanism (3) to the second roller shaft (12).

7. The laser vision-based bar and wire rolling force measuring device according to claim 6, characterized in that, The visual recognition unit includes a laser (51), a camera (52), and a calibration plate (53). The laser (51) is used to emit laser light to the calibration plate (53). The rolled bar wire is located between the laser (51) and the calibration plate (53). The camera (52) is used to capture the contour image on the calibration plate (53).

8. The laser vision-based bar and wire rolling force measuring device according to claim 7, characterized in that, The laser propagation direction of the laser (51) is perpendicular to the calibration plate (53); The camera (52) is angled to photograph the calibration plate (53).

9. The laser vision-based bar and wire rolling force measuring device according to claim 6, characterized in that, The data acquisition unit includes an oil pressure sensor (61) and a pressure sensor (62). The oil pressure sensor (61) is used to detect the oil pressure of the oil film bearing (4), and the pressure sensor (62) is used to monitor the pressure output by the pressurizing mechanism (3).

Images