Automatic feeding control method and device for mine rough crushing, electronic equipment and storage medium

Abstract

The invention provides an automatic feeding control method and device for mine rough crushing, electronic equipment and a storage medium, and the automatic feeding control method for mine rough crushing comprises the steps that operation data and machine vision recognition data of a rough crushing system are obtained, the machine vision identification data comprises material form detection data, crushing hopper material quantity detection data, signal bin material detection data and foreign matter detection data; and based on the operation data, the material form detection data, the crushing hopper material quantity detection data, the signal bin material detection data and the foreign matter detection data, feeding control data are determined, and the feeding control data comprise the feeding speed. According to the method, automatic feeding of mine rough crushing can be achieved in combination with the operation data and the machine vision recognition data, manual operation is not needed, and therefore the feeding speed control precision is improved, and manual dependence is reduced.

Description

Technical Field

[0001] The present application relates to the field of coarse crushing technology in mines, and in particular to a method, device, electronic equipment and storage medium for automatic control of feed in coarse crushing in mines. Background Art

[0002] Currently, the feed rate for coarse crushing in mines is traditionally controlled manually. This method suffers from low adjustment precision and repetitive labor. Furthermore, manual judgment is also affected by personnel quality, making it impossible to achieve optimal throughput for the crushing equipment. Therefore, the existing coarse crushing feed control method relies on manual operation, resulting in low feed rate control precision and a waste of human resources. Summary of the Invention

[0003] The purpose of the embodiments of the present application is to provide a method, device, electronic device and storage medium for automatic control of feeding of coarse crushing in mines, which are used to combine operating data and machine vision recognition data to realize automatic feeding of coarse crushing in mines without manual operation, thereby improving the feeding speed control accuracy and reducing manual dependence.

[0004] In a first aspect, the present invention provides a method for automatically controlling feed of a coarse crusher in a mine, the method comprising: Acquire the operation data and machine vision recognition data of the coarse crushing system, wherein the machine vision recognition data includes material shape detection data, crushing bucket material quantity detection data, signal bin material detection data, and foreign matter detection data; Based on the operation data, the material form detection data, the crushing bucket material quantity detection data, the signal bin material detection data and the foreign matter detection data, feed control data is determined, wherein the feed control data includes a feed speed.

[0005] The method of the first aspect of the present application can obtain the operating data and machine vision recognition data of the coarse crushing system, wherein the machine vision recognition data includes material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, and then can determine the feed control data based on the material shape detection data, the crushing bucket material quantity detection data, the signal bin material detection data and the foreign matter detection data, wherein the feed control data includes the feed speed, and finally, can combine the operating data and the machine vision recognition data to realize automatic feeding of coarse crushing in mines without manual operation, thereby improving the feed speed control accuracy and reducing manual dependence.

[0006] In an optional embodiment, the method further comprises: Get surveillance video stream; Extracting target video frame images based on the monitoring video stream, wherein the target video frame images include a bar screen video frame image, a crushing bucket video frame image, and a signal bin video frame image; Determining the material form detection data based on the bar screen video frame image or the crushing bucket video frame image; Determining the crushing bucket material quantity detection data based on the crushing bucket video frame image; The signal warehouse material detection data is determined based on the signal warehouse video frame picture.

[0007] This optional implementation method obtains a monitoring video stream, and then can extract target video frame images based on the monitoring video stream. The target video frame images include rod screen video frame images, crushing bucket video frame images and signal bin video frame images, and then can determine the material form detection data based on the rod screen video frame images or the crushing bucket video frame images, and then can determine the crushing bucket material quantity detection data based on the crushing bucket video frame images, and thus can determine the signal bin material detection data based on the signal bin video frame images.

[0008] In an optional embodiment, the method further comprises: Processing the target video frame image based on the foreign object detection network to determine the foreign object rectangular frame recognition result; Counting the foreign body rectangular frame recognition results and determining the number of foreign body rectangular frames; The foreign object detection data is determined based on the number of the foreign object rectangular frames.

[0009] This optional implementation can process the target video frame image based on the foreign object detection network to determine the foreign object rectangular frame recognition result, and then by counting the foreign object rectangular frame recognition results and determining the number of foreign object rectangular frames, the foreign object detection data can be determined based on the number of foreign object rectangular frames.

[0010] In an optional embodiment, determining the material form detection data based on the bar screen video frame image or the crushing bucket video frame image includes: Processing the bar screen video frame image or the crushing bucket video frame image based on the first target detection model to determine a material rectangular frame; Determine the long side length of the material rectangular frame; Determining the long side length of the mineral material based on the long side length of the material rectangular frame; The material form detection data is determined based on the long side length of the mineral material and a reference value of the long side length of the mineral material.

[0011] This optional implementation can process the rod screen video frame image or the crushing bucket video frame image based on the first target detection model to determine the material rectangular frame, and then determine the long side length of the material rectangular frame, and then determine the long side length of the mineral material based on the long side length of the material rectangular frame, so as to determine the material morphology detection data based on the long side length of the mineral material and the reference value of the long side length of the mineral material.

[0012] In an optional embodiment, determining the signal warehouse material detection data based on the signal warehouse video frame picture includes: Processing the signal bin video frame image based on a monocular depth estimation model to determine a depth map of the signal bin video frame image; determining an average inverse depth value based on the depth map; The signal bin material detection data is determined based on the average inverse depth value and the target fitting curve.

[0013] This optional implementation can process the signal warehouse video frame image based on a monocular depth estimation model to determine a depth map of the signal warehouse video frame image, wherein the depth map includes the inverse depth value of each pixel point in the signal warehouse video frame image, and then the average inverse depth value can be determined based on the depth map, thereby determining the signal warehouse material detection data based on the average inverse depth value and the target fitting curve.

[0014] In an optional embodiment, the determining the crushing bucket material quantity detection data based on the crushing bucket video frame image includes: Processing the crushing bucket video frame image based on the second target detection model to obtain crushing bucket rectangular frame data; Inputting the crushing bucket rectangular frame data and the crushing bucket video frame image into a non-semantic segmentation network to obtain a segmentation result of the mineral material inside the crushing bucket based on the non-semantic segmentation network; Determine an inner area of the crushing bucket, and determine a weight of each pixel in the inner area of the crushing bucket; The crushing bucket material amount detection data is determined based on the weight of each pixel in the inner area of the crushing bucket and the segmentation result.

[0015] This optional implementation manner can process the crushing bucket video frame image based on the second target detection model to obtain crushing bucket rectangular frame data, and then can input the crushing bucket rectangular frame data and the crushing bucket video frame image into a non-semantic segmentation network to obtain a segmentation result of the mineral material inside the crushing bucket based on the non-semantic segmentation network, and then can determine the internal area of the crushing bucket and determine the weight of each pixel in the internal area of the crushing bucket, so as to determine the crushing bucket material quantity detection data based on the weight of each pixel in the internal area of the crushing bucket and the segmentation result.

[0016] In an optional embodiment, the determining of the feed control data based on the operation data, the material form detection data, the crushing bucket material quantity detection data, the signal bin material detection data, and the foreign matter detection data includes: Determine the status of the jaw crusher and the current belt ore load based on the operating data; Determining a feed rate based on the state of the jaw crusher and the current belt ore load; The feed speed is corrected based on the material form detection data, the crushing bucket material amount detection data, the signal bin material detection data and the foreign matter detection data.

[0017] This optional implementation can determine the state of the jaw crusher and the current belt ore load based on the operating data, and further can determine the feed speed based on the state of the jaw crusher and the current belt ore load, thereby being able to correct the feed speed based on the material form detection data, the crushing bucket material load detection data, the signal bin material detection data and the foreign matter detection data.

[0018] In an optional embodiment, the correcting the feed speed based on the material form detection data, the crushing bucket material amount detection data, the signal bin material detection data and the foreign matter detection data includes: When the foreign matter detection data indicates that a foreign matter has been detected, feeding is stopped; When the signal bin material detection data indicates that the signal bin is empty, the feeding is stopped; When the operating data of the coarse crushing system indicates that the current of the jaw crusher is abnormal, the feeding is stopped.

[0019] This optional implementation method can stop feeding when foreign matter is detected, when the signal bin is empty, or when the current of the jaw crusher is abnormal, thereby preventing accidents.

[0020] In a second aspect, the present invention provides an automatic feed control device for a coarse crusher in a mine, the device comprising: An acquisition module is used to acquire the operation data of the coarse crushing system and machine vision recognition data, wherein the machine vision recognition data includes material form detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data; A control module is used to determine feed control data based on the operating data, the material form detection data, the crushing bucket material quantity detection data, the signal bin material detection data and the foreign matter detection data, wherein the feed control data includes a feed speed.

[0021] The device of the second aspect of the present application is capable of obtaining the operating data and machine vision recognition data of the coarse crushing system, wherein the machine vision recognition data includes material form detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, and then can determine the feed control data based on the material form detection data, the crushing bucket material quantity detection data, the signal bin material detection data and the foreign matter detection data, wherein the feed control data includes the feed speed, and finally, can combine the operating data and the machine vision recognition data to realize automatic feeding of coarse crushing in mines without manual operation, thereby improving the feed speed control accuracy and reducing manual dependence.

[0022] In a third aspect, the present invention provides an electronic device, comprising: processor; and The memory is configured to store machine-readable instructions, which, when executed by the processor, execute the automatic feeding control method for coarse crushing in a mine as described in any one of the aforementioned embodiments.

[0023] The electronic device of the third aspect of the present application can obtain the operating data and machine vision recognition data of the coarse crushing system by executing the automatic control method of feeding of coarse crushing in mines, wherein the machine vision recognition data includes material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, and then can determine the feeding control data based on the material shape detection data, the crushing bucket material quantity detection data, the signal bin material detection data and the foreign matter detection data, wherein the feeding control data includes the feeding speed, and finally, can combine the operating data and the machine vision recognition data to realize automatic feeding of coarse crushing in mines without manual operation, thereby improving the accuracy of feeding speed control and reducing manual dependence.

[0024] In a fourth aspect, the present invention provides a storage medium storing a computer program, wherein the computer program is executed by a processor to implement the method for automatically controlling feed of a coarse crusher in a mine as described in any one of the aforementioned embodiments.

[0025] The storage medium of the fourth aspect of the present application can obtain the operation data and machine vision recognition data of the coarse crushing system by executing the automatic control method of feeding of coarse crushing in mines, wherein the machine vision recognition data includes material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, and then can determine the feeding control data based on the material shape detection data, the crushing bucket material quantity detection data, the signal bin material detection data and the foreign matter detection data, wherein the feeding control data includes the feeding speed, and finally, can combine the operation data and the machine vision recognition data to realize automatic feeding of coarse crushing in mines without manual operation, thereby improving the accuracy of feeding speed control and reducing manual dependence. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without creative work.

[0027] Figure 1 This is a flow chart of an automatic control method for feeding coarse crushing in a mine disclosed in an embodiment of the present application; Figure 2 This is a schematic structural diagram of an automatic feed control device for coarse crushing in a mine disclosed in an embodiment of the present application; Figure 3 This is a schematic structural diagram of an electronic device disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0028] The technical solutions in the embodiments of the present application will be described below in conjunction with the drawings in the embodiments of the present application.

[0029] Currently, the feed rate for coarse crushing in mines is traditionally controlled manually. This method suffers from low adjustment precision and repetitive labor. Furthermore, manual judgment is also affected by personnel quality, making it impossible to achieve optimal throughput for the crushing equipment. Therefore, the existing coarse crushing feed control method relies on manual operation, resulting in low feed rate control precision and a waste of human resources.

[0030] In response to the technical defects of the existing technology, the embodiments of the present application provide a method, device, electronic equipment and storage medium for automatic feeding control of coarse crushing in mines, which can obtain the operating data and machine vision recognition data of the coarse crushing system, wherein the machine vision recognition data includes material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, and then can determine the feeding control data based on the material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, wherein the feeding control data includes the feeding speed, and finally, can combine the operating data and machine vision recognition data to realize automatic feeding of coarse crushing in mines without manual operation, thereby improving the feeding speed control accuracy and reducing manual dependence.

[0031] See also Figure 1 , Figure 1 This is a flow chart of an automatic control method for feeding coarse crushing in a mine disclosed in an embodiment of the present application. Figure 1 As shown, the feeding automatic control method of the mine coarse crushing disclosed in the embodiment of the present application includes the following steps: 101. Obtaining the operating data and machine vision recognition data of the coarse crushing system, wherein the machine vision recognition data includes material shape detection data, crushing bucket material quantity detection data, signal bin material detection data, and foreign matter detection data; 102. Determine feed control data based on operation data, material form detection data, crushing bucket material quantity detection data, signal bin material detection data, and foreign matter detection data, wherein the feed control data includes feed speed.

[0032] The embodiment of the present application can obtain the operating data and machine vision recognition data of the coarse crushing system, wherein the machine vision recognition data includes material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, and then can determine the feed control data based on the material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, wherein the feed control data includes the feed speed, and finally, can combine the operating data and machine vision recognition data to realize automatic feeding of coarse crushing in mines without manual operation, thereby improving the feed speed control accuracy and reducing manual dependence.

[0033] In an embodiment of the present application, the coarse crushing system includes a mineral crushing device and a controller for controlling the mineral crushing device, as well as various sensors, wherein the controller can be a PLC controller that can generate PLC data based on data detected by the sensors.

[0034] In the embodiment of the present application, the operating data of the coarse crushing system refers to the data generated by the coarse crushing system during production operation, which includes the real-time current of the jaw crusher, the current belt ore volume, and the signal bin indicator light information.

[0035] In an embodiment of the present application, machine vision recognition data is data obtained by processing a video or picture based on a machine vision algorithm, for example, data obtained by processing a video data stream based on an image processing algorithm.

[0036] In the embodiment of the present application, in order to obtain the input data required by the machine vision algorithm, the coarse crushing system also includes a plurality of imaging devices, which are used to capture real-time images of the areas where the various mineral crushing equipment are located.

[0037] In the embodiment of the present application, the feed control data includes the feed speed, wherein the feed speed can be controlled to fully utilize the production capacity of the mineral crushing equipment, and to avoid abnormal operation of the mineral crushing equipment due to excessive mineral materials.

[0038] In the embodiment of the present application, the material form detection data refers to data that characterizes the recognition result of the material form. For example, the material form detection data can characterize that the current belt is transporting oversized mineral materials.

[0039] In the embodiment of the present application, the crushing bucket material quantity detection data refers to data representing the result of identifying the amount of mineral material in the crushing bucket. For example, the crushing bucket material quantity detection data may represent that the crushing bucket material quantity is 50%.

[0040] In the embodiment of the present application, the signal bin material detection data refers to data representing the identification result of the remaining content of mineral materials in the signal bin. For example, the signal bin material detection number can represent that the remaining content of mineral materials in the signal bin is 50%.

[0041] In the embodiment of the present application, the foreign matter detection data refers to data representing the identification results of large foreign matter. For example, the foreign matter detection data may represent that there is non-mineral material on the belt with a diameter exceeding a preset value.

[0042] In an embodiment of the present application, the feed control data also includes the vibration frequency of the bar screen, wherein the higher the vibration frequency of the bar screen, the faster the crushed mineral material is output, thereby allowing the belt to carry more mineral material within a specified time.

[0043] In the embodiment of the present application, as an optional implementation, the method for automatically controlling the feeding of coarse crushing in a mine further includes the following steps: Get surveillance video stream; Extract target video frame images based on the monitoring video stream, the target video frame images include bar screen video frame images, crushing bucket video frame images and signal bin video frame images; Determine material shape detection data based on the video frame images of the bar screen or the crushing bucket; Determine the crushing bucket material quantity detection data based on the crushing bucket video frame image; Determine the signal warehouse material detection data based on the signal warehouse video frame image.

[0044] This optional implementation method obtains the monitoring video stream, and then can extract target video frame images based on the monitoring video stream. The target video frame images include rod screen video frame images, crushing bucket video frame images and signal bin video frame images. Then, based on the rod screen video frame images or the crushing bucket video frame images, the material form detection data can be determined, and then based on the crushing bucket video frame images, the crushing bucket material quantity detection data can be determined, and thus the signal bin material detection data can be determined based on the signal bin video frame images.

[0045] In this optional embodiment, the monitoring video stream is generated by an imaging device, for example, it can be generated by a monitor.

[0046] In this optional embodiment, the surveillance video stream is composed of multiple frames of images, namely, multiple video frame images. Since the imaging device captures real-time images of the device, these real-time images include images of the device in operation and images of the device in standby mode. Therefore, it is necessary to extract images of the device in operation from the surveillance video stream in order to analyze the feed rate required for the device to operate based on these images of the device in operation. Therefore, the target video frame image refers to an image that reflects the operation of the device.

[0047] In this optional real-time mode, the bar screen video frame picture refers to an image reflecting the operation of the bar screen, the crushing bucket video frame picture refers to an image reflecting the operation of the crushing bucket, and the signal bin video frame picture refers to an image reflecting the operation of the signal bin.

[0048] In the embodiment of the present application, as an optional implementation, the method for automatically controlling the feeding of coarse crushing in a mine further includes the following steps: Process the target video frame image based on the foreign object detection network to determine the foreign object rectangular frame recognition result; Counting the foreign body rectangular frame recognition results and determining the number of foreign body rectangular frames; The foreign object detection data is determined based on the number of foreign object rectangular frames.

[0049] This optional implementation can process the target video frame image based on the foreign object detection network to determine the foreign object rectangular frame recognition result, and then by counting the foreign object rectangular frame recognition results and determining the number of foreign object rectangular frames, it can determine the foreign object detection data based on the number of foreign object rectangular frames.

[0050] In this optional embodiment, the foreign object detection network is a deep learning model, wherein the deep learning model is trained through training data.

[0051] In this optional embodiment, processing the target video frame image based on the foreign object detection network means using the target video frame image as input data for the foreign object detection network. Accordingly, the foreign object detection network processes the input data based on pre-trained parameters and algorithms, specifically identifying foreign objects in the target video frame image and framing them with a foreign object rectangular frame.

[0052] In this optional embodiment, multiple foreign objects can be recorded in a target video frame image. Therefore, the number of foreign object rectangular boxes can be multiple, and the foreign object detection data can be determined based on the number of foreign object rectangular boxes. That is, it can be determined whether there are foreign objects based on the number of foreign object rectangular boxes. For example, when the number of foreign object rectangular boxes is 0, it can be determined that there are no foreign objects, and if the number of foreign object rectangular boxes is greater than 0, it can be determined that there are foreign objects.

[0053] In the embodiment of the present application, as an optional implementation, the step of determining the material form detection data based on the bar screen video frame image or the crushing bucket video frame image includes the following sub-steps: Processing the bar screen video frame image or the crushing bucket video frame image based on the first target detection model to determine the material rectangular frame; Determine the long side length of the material rectangle; Determine the long side length of the mineral material based on the long side length of the material rectangle; The material shape detection data is determined based on the long side length of the mineral material and the reference value of the long side length of the mineral material.

[0054] This optional implementation can process the video frame image of the rod screen or the video frame image of the crushing bucket based on the first target detection model to determine the material rectangular frame, and then determine the long side length of the material rectangular frame, and then determine the long side length of the mineral material based on the long side length of the material rectangular frame, so as to determine the material morphology detection data based on the long side length of the mineral material and the reference value of the long side length of the mineral material.

[0055] In this optional embodiment, processing the grating screen video frame image or the crushing bucket video frame image based on the first target detection model means that the crushing bucket video frame image or the grating screen video frame image can be used as input data of the first target detection model.

[0056] In this optional embodiment, the first target detection model can be obtained by training the bar screen video training data, wherein the first target detection model is used to identify whether there is mineral material in the bar screen video frame image or the crushing bucket video frame image, wherein the mineral material can be divided into extra-large blocks, large blocks and other specifications of mineral material according to the diameter of the mineral material.

[0057] In this optional embodiment, the long side length of the material rectangular frame refers to the length of the longest side of the material rectangular frame.

[0058] In this optional embodiment, the reference value of the long side length of the mineral material includes reference values of the long side length of mineral materials of various sizes and specifications. A specific method of determining the material morphology detection data based on the long side length of the mineral material and the reference value of the long side length of the mineral material is: The long side length of the mineral material is compared with the long side length reference values of mineral materials of various sizes and specifications in sequence, and the material form detection data is determined based on the comparison results.

[0059] In the embodiment of the present application, as an optional implementation, the step of determining the signal bin material detection data based on the signal bin video frame image includes the following steps: Processing the signal bin video frame image based on the monocular depth estimation model to determine a depth map of the signal bin video frame image, the depth map including an inverse depth value of each pixel point in the signal bin video frame image; determining an average inverse depth value based on the depth map; The signal bin material detection data is determined based on the average inverse depth value and the target fitting curve.

[0060] This optional implementation can process the signal warehouse video frame image based on the monocular depth estimation model to determine the depth map of the signal warehouse video frame image. The depth map includes the inverse depth value of each pixel point in the signal warehouse video frame image, and then the average inverse depth value can be determined based on the depth map, so that the signal warehouse material detection data can be determined based on the average inverse depth value and the target fitting curve.

[0061] In this optional embodiment, the monocular depth estimation model refers to a model that predicts the depth information of each pixel from a single RGB image.

[0062] In this optional implementation, determining the average inverse depth value based on the depth map refers to calculating the average of the inverse depth values of each pixel in the signal bin video frame image, thereby obtaining the average inverse depth value.

[0063] In this optional embodiment, the target fitting curve is pre-constructed, wherein the target fitting curve reflects the correspondence between the inverse depth value and the volume ratio of the mineral material in the signal bin. Therefore, after determining the average inverse depth value, the volume ratio of the mineral material in the signal bin corresponding to the average inverse depth value can be determined according to the target fitting curve, thereby determining the signal bin material detection data.

[0064] In the embodiment of the present application, as an optional implementation, the step of determining the crushing bucket material quantity detection data based on the crushing bucket video frame image includes the following steps: Processing the crushing bucket video frame image based on the second target detection model to obtain crushing bucket rectangular frame data; Input the crushing bucket rectangular frame data and the crushing bucket video frame image into the non-semantic segmentation network to obtain the segmentation result of the mineral material inside the crushing bucket based on the non-semantic segmentation network; Determine the inner area of the crushing bucket and determine the weight of each pixel in the inner area of the crushing bucket; Crushing bucket material amount detection data is determined based on the weight of each pixel in the inner area of the crushing bucket and the segmentation result.

[0065] This optional implementation method can process the crushing bucket video frame image based on the second target detection model to obtain the crushing bucket rectangular frame data, and then can input the crushing bucket rectangular frame data and the crushing bucket video frame image into the non-semantic segmentation network to obtain the segmentation result of the mineral material inside the crushing bucket based on the non-semantic segmentation network, and then can determine the internal area of the crushing bucket and determine the weight of each pixel in the internal area of the crushing bucket, so as to determine the crushing bucket material quantity detection data based on the weight of each pixel in the internal area of the crushing bucket and the segmentation result.

[0066] In this optional embodiment, the second object detection model is trained based on crushing bucket image training data, and the second object detection model is used to identify the area containing the crushing bucket in the image.

[0067] In this optional embodiment, the crushing bucket rectangular frame data may refer to the coordinates of the crushing bucket rectangular frame.

[0068] In this optional embodiment, the non-semantic segmentation network first crops out the area marked by the crushing bucket rectangular box from the crushing bucket video frame image based on the crushing bucket rectangular box, that is, crops out the image area containing the crushing bucket, and then marks the sub-area where the mineral material exists from the image containing the crushing bucket and segments the sub-area.

[0069] In this optional embodiment, the inner area of the crushing bucket may refer to the area within the crushing bucket containing the ore. The inner area of the crushing bucket may be defined by the crushing bucket rectangular frame. For example, the crushing bucket rectangular frame may be reduced based on the wall thickness of the crushing bucket to obtain an image area marked by the reduced crushing bucket rectangular frame.

[0070] In this optional embodiment, since the crushing bucket is usually conical, the degree to which each pixel of the image of the internal area of the crushing bucket reflects the volume is different. For example, when the pixel in the middle of the image is the same as the pixel at the edge of the image, since the pixel in the middle of the image is a two-dimensional image of the middle of the crushing bucket, the depth of its reflection is large, and thus the volume of the mineral material is large, that is, the weight is high.

[0071] In this optional embodiment, determining the crushing bucket material quantity detection data based on the weight of each pixel in the internal area of the crushing bucket and the segmentation result means multiplying the weight of each pixel in the internal area of the crushing bucket with the segmentation result to obtain the crushing bucket material quantity detection data.

[0072] In the embodiment of the present application, the feed control data is determined based on the operation data, the material form detection data, the crushing bucket material amount detection data, the signal bin material detection data and the foreign matter detection data. A specific manner in which the feed control data includes the feed speed is: Determine the status of the jaw crusher and the current belt ore load based on operating data; Determine the feed rate based on the status of the jaw crusher and the current belt load; The feed speed is corrected based on the material shape detection data, the crushing bucket material quantity detection data, the signal bin material detection data and the foreign matter detection data.

[0073] This optional implementation method can determine the status of the jaw crusher and the current belt ore load based on the operating data, and then determine the feed speed based on the status of the jaw crusher and the current belt ore load, and then correct the feed speed based on the material shape detection data, the crushing bucket material load detection data, the signal bin material detection data and the foreign matter detection data.

[0074] In this optional embodiment, the current belt mineral content refers to the mineral content on the belt.

[0075] In this optional implementation manner, the state of the jaw crusher may refer to the current state of the jaw crusher.

[0076] In the embodiment of the present application, as an optional implementation, the method for automatically controlling the feeding of coarse crushing in a mine further includes the following steps: When the foreign matter detection data indicates that a foreign matter has been detected, the feeding is stopped; When the signal bin material detection data indicates that the signal bin is empty, the feeding is stopped; When the operating data of the coarse crushing system indicates that the current of the jaw crusher is abnormal, the feeding is stopped.

[0077] This optional implementation method can stop feeding when foreign matter is detected, when the signal bin is empty, or when the current of the jaw crusher is abnormal, thereby preventing accidents.

[0078] In this optional embodiment, if the foreign object detection data indicates that a foreign object has been detected, a foreign object alarm may be generated to prompt relevant personnel to remove the foreign object in a timely manner.

[0079] In this optional implementation manner, if the jaw crusher current is abnormal, a jaw crusher abnormality alarm may be generated, and the feeding may be resumed after the jaw crusher current returns to normal.

[0080] In an embodiment of the present application, the state of the signal bin indicator light can also be controlled according to the remaining material amount in the signal bin. For example, when the remaining material amount in the signal bin indicates that the signal bin is about to be empty, the signal bin indicator light is controlled to display green to prompt relevant personnel to replenish the signal bin with mineral materials in time. If the remaining material amount in the signal bin indicates that the signal bin is about to be full, the signal bin indicator light is controlled to display red to prompt relevant personnel to stop replenishing the signal bin with mineral materials in time.

[0081] See also Figure 2 , Figure 2 This is a schematic diagram of the structure of an automatic feed control device for a coarse crushing mine disclosed in an embodiment of the present application. Figure 2 As shown, the feeding automatic control device of the mine coarse crushing embodiment of the present application includes the following functional modules: An acquisition module 201 is used to acquire the operation data of the coarse crushing system and machine vision recognition data, wherein the machine vision recognition data includes material form detection data, crushing bucket material quantity detection data, signal bin material detection data, and foreign matter detection data; The control module 202 is used to determine the feed control data based on the operation data, the material form detection data, the crushing bucket material quantity detection data, the signal bin material detection data and the foreign matter detection data, wherein the feed control data includes the feed speed.

[0082] The device of the embodiment of the present application can obtain the operating data and machine vision recognition data of the coarse crushing system, wherein the machine vision recognition data includes material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, and then can determine the feed control data based on the material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, wherein the feed control data includes the feed speed, and finally, can combine the operating data and machine vision recognition data to realize automatic feeding of coarse crushing in mines without manual operation, thereby improving the feed speed control accuracy and reducing manual dependence.

[0083] See also Figure 3 , Figure 3 This is a schematic diagram of the structure of an electronic device disclosed in an embodiment of the present application. Figure 3 As shown, the feeding automatic control device of the mine coarse crushing embodiment of the present application includes the following functional modules: Processor 301; and The memory 302 is configured to store machine-readable instructions. When the instructions are executed by the processor 301, the method for automatically controlling feed of rough crushing in a mine as described in any of the aforementioned embodiments is executed.

[0084] The electronic device of the embodiment of the present application can obtain the operating data and machine vision recognition data of the coarse crushing system by executing the automatic control method of feeding of coarse crushing in mines, wherein the machine vision recognition data includes material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, and then can determine the feeding control data based on the material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, wherein the feeding control data includes the feeding speed, and finally, can combine the operating data and the machine vision recognition data to realize automatic feeding of coarse crushing in mines without manual operation, thereby improving the accuracy of feeding speed control and reducing manual dependence.

[0085] An embodiment of the present application provides a storage medium storing a computer program, and the computer program is executed by a processor as the automatic feeding control method for coarse crushing in a mine according to any of the aforementioned embodiments.

[0086] The storage medium of the embodiment of the present application can obtain the operation data and machine vision recognition data of the coarse crushing system by executing the automatic control method of feeding of coarse crushing in mines, wherein the machine vision recognition data includes material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, and then can determine the feeding control data based on the material shape detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data, wherein the feeding control data includes the feeding speed, and finally, can combine the operation data and the machine vision recognition data to realize automatic feeding of coarse crushing in mines without manual operation, thereby improving the accuracy of feeding speed control and reducing manual dependence.

[0087] In the embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interface, the indirect coupling or communication connection of the device or unit can be electrical, mechanical or other forms.

[0088] In addition, the units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0089] Furthermore, the functional modules in each embodiment of the present application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0090] It should be noted that if the function is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the existing technology, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program code.

[0091] In this document, relational terms such as first and second, etc. are used merely to distinguish one entity or operation from another entity or operation, but do not necessarily require or imply any actual relationship or order between these entities or operations.

[0092] The above are merely examples of the present application and are not intended to limit the scope of protection of the present application. Those skilled in the art will appreciate that various modifications and variations are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application shall be included within the scope of protection of the present application.

Claims

1. A method for automatically controlling the feeding of coarse crushing in a mine, characterized in that: The method comprises: Acquire the operation data and machine vision recognition data of the coarse crushing system, wherein the machine vision recognition data includes material shape detection data, crushing bucket material quantity detection data, signal bin material detection data, and foreign matter detection data; Based on the operation data, the material form detection data, the crushing bucket material quantity detection data, the signal bin material detection data and the foreign matter detection data, feed control data is determined, wherein the feed control data includes a feed speed.

2. The method according to claim 1, wherein The method further comprises: Get surveillance video stream; Extracting target video frame images based on the monitoring video stream, wherein the target video frame images include a bar screen video frame image, a crushing bucket video frame image, and a signal bin video frame image; Determining the material form detection data based on the bar screen video frame image or the crushing bucket video frame image; Determining the crushing bucket material quantity detection data based on the crushing bucket video frame image; The signal warehouse material detection data is determined based on the signal warehouse video frame picture.

3. The method according to claim 2, wherein The method further comprises: Processing the target video frame image based on the foreign object detection network to determine the foreign object rectangular frame recognition result; Counting the foreign body rectangular frame recognition results and determining the number of foreign body rectangular frames; The foreign object detection data is determined based on the number of the foreign object rectangular frames.

4. The method according to claim 2, wherein The determining of the material form detection data based on the bar screen video frame image or the crushing bucket video frame image includes: Processing the bar screen video frame image or the crushing bucket video frame image based on the first target detection model to determine a material rectangular frame; Determine the long side length of the material rectangular frame; Determining the long side length of the mineral material based on the long side length of the material rectangular frame; The material form detection data is determined based on the long side length of the mineral material and a reference value of the long side length of the mineral material.

5. The method according to claim 2, wherein The determining of the signal warehouse material detection data based on the signal warehouse video frame picture includes: Processing the signal bin video frame image based on a monocular depth estimation model to determine a depth map of the signal bin video frame image; determining an average inverse depth value based on the depth map; The signal bin material detection data is determined based on the average inverse depth value and the target fitting curve.

6. The method according to claim 2, wherein The determining of the crushing bucket material quantity detection data based on the crushing bucket video frame image includes: Processing the crushing bucket video frame image based on the second target detection model to obtain crushing bucket rectangular frame data; Inputting the crushing bucket rectangular frame data and the crushing bucket video frame image into a non-semantic segmentation network to obtain a segmentation result of the mineral material inside the crushing bucket based on the non-semantic segmentation network; Determine an inner area of the crushing bucket, and determine a weight of each pixel in the inner area of the crushing bucket; The crushing bucket material amount detection data is determined based on the weight of each pixel in the inner area of the crushing bucket and the segmentation result.

7. The method according to claim 1, wherein The determining of the feed control data based on the operation data, the material form detection data, the crushing bucket material quantity detection data, the signal bin material detection data and the foreign matter detection data includes: Determine the status of the jaw crusher and the current belt ore load based on the operating data; Determining a feed rate based on the state of the jaw crusher and the current belt ore load; The feed speed is corrected based on the material form detection data, the crushing bucket material amount detection data, the signal bin material detection data and the foreign matter detection data.

8. The method according to claim 7, wherein The correction of the feeding speed based on the material form detection data, the crushing bucket material amount detection data, the signal bin material detection data and the foreign matter detection data includes: When the foreign matter detection data indicates that a foreign matter has been detected, feeding is stopped; When the signal bin material detection data indicates that the signal bin is empty, the feeding is stopped; When the operating data of the coarse crushing system indicates that the current of the jaw crusher is abnormal, the feeding is stopped.

9. An automatic feeding control device for coarse crushing in mines, characterized in that: The device comprises: An acquisition module is used to acquire the operation data of the coarse crushing system and machine vision recognition data, wherein the machine vision recognition data includes material form detection data, crushing bucket material quantity detection data, signal bin material detection data and foreign matter detection data; A control module is used to determine feed control data based on the operating data, the material form detection data, the crushing bucket material quantity detection data, the signal bin material detection data and the foreign matter detection data, wherein the feed control data includes a feed speed.

10. An electronic device, characterized in that: include: processor; as well as A memory configured to store machine-readable instructions, which, when executed by the processor, execute the automatic feeding control method for coarse crushing in a mine as described in any one of claims 1 to 8.

11. A storage medium, characterized in that: The storage medium stores a computer program, and the computer program is executed by a processor to implement the automatic feeding control method for coarse crushing in a mine according to any one of claims 1 to 8.

Images