Online analysis method for dumping granularity of mining truck

Abstract

An on-line analysis method for the dumping granularity of a mining truck comprises the steps that one or more cameras are installed above or obliquely above the dumping position of the mining truck, data transmission connection is established between the cameras and a central control room work station, and before on-line analysis is started, the cameras are connected with the central control room work station; the method comprises the following steps: firstly, obtaining a proportional relation among a shuffled ore area, a pixel size and a real object size through an offline shuffled image, and training an ore recognition model; during online analysis, cutting a real-time shuffling image according to a shuffling ore area, inputting the real-time shuffling image into an ore recognition model, outputting pixel shape detection information of each ore, and then performing granularity calculation on the image when the mining truck is in a normal shuffling state to obtain a real particle size length of each ore; then, the number of pixels in each piece of ore is fitted, and the proportion of ores of different types and fractions is counted according to the real particle size length of each piece of ore. According to the method, the ore granularity in the dumping ore card can be accurately analyzed online in real time under the complex background.

Description

Technical Field

[0001] The invention relates to the field of intelligent control methods for dumping materials by mining trucks, and in particular to an online analysis method for particle size of dumped materials by mining trucks. Background Art

[0002] In mining production, evaluating the blasting effect of raw ore and real-time monitoring of crusher efficiency are crucial to ensure the efficient operation of the production chain. The feed particle size distribution of the primary crushing station is a key indicator for evaluating the blasting effect and crusher efficiency. Unsatisfactory blasting will cause crusher blockage and shutdown, which not only reduces the service life of the equipment, but also hinders the realization of production goals, thereby affecting the efficiency of the entire production line. In order to improve the working efficiency of the crusher and reduce energy consumption, precise control of feed particle size becomes particularly critical.

[0003] The traditional manual screening method is not only inefficient and costly, but also labor-intensive and has prominent safety hazards. In addition, when dealing with raw ores with larger particle sizes, the manual identification method has low accuracy and is not timely. The online monitoring method is difficult to implement. Since the unloading of the mine truck is a dynamic and changeable process, the continuous change of the position of the mine truck bucket causes the ratio between the pixel size and the physical size in the image to change continuously, which increases the calculation complexity. In addition, since the mine truck bucket is generally designed to be deep, it is difficult to directly observe the ore inside, which further increases the difficulty of accurately calculating the ore particle size distribution. Therefore, a method is needed to accurately analyze the particle size of the ore unloading mine truck in real time under complex backgrounds. Summary of the invention

[0004] The purpose of the present invention is to provide an online analysis method for the particle size of material dumped by a mining truck, which can accurately and real-time analyze the particle size of ore in the material dumped by the mining truck under complex background.

[0005] The technical solution adopted by the present invention to solve the above technical problems is: an online analysis method for the particle size of material discharged by a mining truck, comprising the following steps: Step 1: Equipment Installation Install one or more cameras above or diagonally above the unloading position of the mining truck, and establish a data transmission connection between the camera and the workstation in the central control room; Step 2: Offline processing 2.1. Before starting the online analysis, the camera continuously collects the dumping images of the mining truck to obtain the offline dumping images within a period of time; 2.2. The mining truck area and the ore dumping area are divided from the offline dumping image. The ore dumping area is the location where the ore in the falling state exists between the mining truck and the ore dumping accumulation position. Then, the proportional relationship between the pixel size in the image and the actual size of the photographed object is obtained based on the size of the mining truck area on the image and the actual size of the mining truck. 2.3. Clean, screen and annotate the offline dumping images to obtain the target detection data set, divide the target detection data set into a training set and a validation set, input the training set into the visual target detection network for training, and evaluate the model with the highest score through the validation set as the ore recognition model; Step 3: Online Analysis 3.1. Start online analysis, continuously collect images of mining truck dumping through cameras, and obtain real-time dumping images; 3.2. The real-time dumping image is cropped according to the dumping ore area, and the cropped real-time image of the dumping area is filtered and denoised; then the real-time image of the dumping area is input into the ore recognition model, the ore in the real-time image of the dumping area is recognized by the ore recognition model, and the pixel shape detection information of each ore is output; 3.3. Identify the mining trucks in the real-time dumping image and classify the mining trucks into the state of not reaching the dumping position, the normal dumping state, the dumping stop state and the empty truck leaving state; When the mining truck is in a normal unloading state, the pixel shape detection information of the ore is used to calculate the particle size according to the proportional relationship between the pixel size and the actual size, and the real particle size length of each ore is obtained. Then, the number of pixels inside each ore is fitted using OpenCV, and the proportion of different types of ore particles is calculated based on the real particle size length of each ore. 3.4. Based on the pixel shape detection information of the ore, each ore is drawn on the real-time dumping image, and the drawn image is displayed on the workstation in the central control room.

[0006] Preferably, each camera is equipped with an LED explosion-proof lamp.

[0007] Preferably, an audible and visual alarm is installed near the unloading position of the mining truck. When the actual particle size length of a certain ore obtained through particle size calculation is greater than the over-limit threshold, the image of the over-limit ore is saved, and an alarm is issued to the mining truck driver through the audible and visual alarm; when the unloading of the mining truck is completed, the maximum ore particle size and the number of over-limit ore in the unloading process are counted.

[0008] Preferably, the workstation in the central control room performs data processing according to the particle size calculation results, and generates a chart showing the proportion of each particle size of the ore and a particle size change curve.

[0009] According to the above technical solution, the beneficial effects of the present invention are: The present invention can accurately and in real time analyze the ore particle size in the unloading ore card online under complex backgrounds, and has the following advantages: 1. Reduce labor costs, and a single workstation can manage the conditions of multiple unloading locations; 2. No manual screening is required, and the particle size distribution data of each car is counted in real time; 3. Based on the particle size distribution data, the crushing effect of the upstream raw ore and the material carrying situation of the ore card are evaluated; 4. Over-limit ore is discovered in time to avoid blockage or damage to the crusher, improve ore processing efficiency, and extend the service life of the crusher; 5. It can assist staff in decision-making, optimize the operating parameters of the crusher, and realize linkage with the intelligent control system. DETAILED DESCRIPTION

[0010] This embodiment is an online analysis method for the particle size of the material dumped by a mining truck, which can accurately and real-time analyze the particle size of the ore in the material dumped by the mining truck under complex background, and specifically includes the following steps: Step 1: Equipment Installation One or more cameras are installed above or diagonally above the unloading position of the mining truck, and a data transmission connection is established between the camera and the workstation in the central control room. Network cable or optical fiber is used to transmit video stream data according to the working distance. When the workstation is close to the camera, a shielded Category 5e network cable is used, and when the workstation is far from the camera, an optical fiber is used. In addition, each camera is equipped with an LED explosion-proof lamp for lighting in dusty and night conditions.

[0011] Step 2: Offline processing 2.1. Before starting the online analysis, the camera continuously collects the unloading images of the mining truck to obtain the offline unloading images within a period of time.

[0012] 2.2. The mining truck area and the ore dumping area are divided from the offline dumping image. The ore dumping area is the location where the ore in the falling state exists between the mining truck and the ore dumping accumulation position.

[0013] Then, a calculation is performed based on the size of the mining truck area on the image and the actual size of the mining truck to obtain a proportional relationship between the pixel size in the image and the actual size of the photographed object.

[0014] 2.3. Clean, screen and annotate the offline dumping images to obtain the target detection data set, divide the target detection data set into a training set and a validation set in a ratio of 8:2, input the training set into the visual target detection network for training, and evaluate the model with the highest score through the validation set as the ore recognition model. It is necessary to continue to collect images with incorrect recognition and add these incorrect recognition images to the data set to optimize the model.

[0015] Step 3: Online Analysis 3.1. Start online analysis and continuously collect images of mining truck dumping through cameras to obtain real-time dumping images.

[0016] 3.2. The real-time dumping image is cropped according to the dumping ore area, and the cropped real-time image of the dumping area is filtered and denoised; then the real-time image of the dumping area is input into the ore recognition model, the ore in the real-time image of the dumping area is recognized by the ore recognition model, and the pixel shape detection information of each ore is output, that is, the shape and size of each ore in the image are obtained.

[0017] 3.3. Identify the mining trucks in the real-time dumping image and divide the mining trucks into the state of not reaching the dumping position, the normal dumping state, the dumping stop state and the empty truck leaving state.

[0018] When the mining truck is in a normal unloading state, the particle size is calculated based on the proportional relationship between the pixel size and the actual size of the ore's pixel shape detection information to obtain the true particle size length of each ore. Then, OpenCV is used to fit the number of pixels inside each piece of ore, and based on the true particle size length of each piece of ore, the proportion of different types of ore particles is counted.

[0019] 3.4. Based on the pixel shape detection information of the ore, each ore is drawn on the real-time dumping image, that is, the position of each ore is framed with a square on the original real-time dumping image, and the drawn image is displayed on the workstation in the central control room.

[0020] The central control room workstation processes data based on the particle size calculation results and generates a chart of the proportion of each particle size of the ore and a particle size change curve, which is updated once each vehicle is unloaded. In addition, users can also query historical particle size data and algorithm output information in the central control room workstation.

[0021] An audible and visual alarm is installed near the unloading position of the mining truck. When the actual particle size length of a certain ore obtained through particle size calculation is greater than the over-limit threshold, the image of the over-limit ore is saved and an alarm is issued to the mining truck driver through the audible and visual alarm. When the mining truck unloads the material, the maximum ore particle size and the number of over-limit ore in the unloading process are counted.

[0022] The above contents are only specific examples and explanations of the concept of the present invention. Any modification, supplement or replacement of the described embodiments by any technician familiar with the technical field shall fall within the protection scope of the present invention as long as it does not deviate from the concept of the invention or exceed the scope defined by the claims.

Claims

1. An online analysis method for the particle size of material discharged from a mining truck, characterized in that: The following steps are involved: Step 1: Equipment Installation Install one or more cameras above or diagonally above the unloading position of the mining truck, and establish a data transmission connection between the camera and the workstation in the central control room; Step 2: Offline processing 2.

1. Before starting the online analysis, the camera continuously collects the dumping images of the mining truck to obtain the offline dumping images within a period of time; 2.

2. The mining truck area and the ore dumping area are divided from the offline dumping image. The ore dumping area is the location where the ore in the falling state exists between the mining truck and the ore dumping accumulation position. Then, the proportional relationship between the pixel size in the image and the actual size of the photographed object is obtained based on the size of the mining truck area on the image and the actual size of the mining truck. 2.

3. Clean, screen and annotate the offline dumping images to obtain the target detection data set, divide the target detection data set into a training set and a validation set, input the training set into the visual target detection network for training, and evaluate the model with the highest score through the validation set as the ore recognition model; Step 3: Online Analysis 3.

1. Start online analysis, continuously collect images of mining truck dumping through cameras, and obtain real-time dumping images; 3.

2. The real-time dumping image is cropped according to the dumping ore area, and the cropped real-time image of the dumping area is filtered and denoised; then the real-time image of the dumping area is input into the ore recognition model, the ore in the real-time image of the dumping area is recognized by the ore recognition model, and the pixel shape detection information of each ore is output; 3.

3. Identify the mining trucks in the real-time dumping image and classify the mining trucks into the state of not reaching the dumping position, the normal dumping state, the dumping stop state and the empty truck leaving state; When the mining truck is in a normal unloading state, the pixel shape detection information of the ore is used to calculate the particle size according to the proportional relationship between the pixel size and the actual size, and the real particle size length of each ore is obtained. Then, the number of pixels inside each ore is fitted using OpenCV, and the proportion of different types of ore particles is calculated based on the real particle size length of each ore. 3.

4. Based on the pixel shape detection information of the ore, each ore is drawn on the real-time dumping image, and the drawn image is displayed on the workstation in the central control room.

2. The online analysis method for the particle size of material discharged from a mining truck according to claim 1 is characterized by: Each camera is equipped with an LED explosion-proof light.

3. The online analysis method for the particle size of material discharged from a mining truck according to claim 1 is characterized by: An audible and visual alarm is installed near the unloading position of the mining truck. When the actual particle size length of a certain ore obtained through particle size calculation is greater than the over-limit threshold, the image of the over-limit ore is saved and an alarm is issued to the mining truck driver through the audible and visual alarm; when the mining truck unloading is completed, the maximum ore particle size and the number of over-limit ore in the unloading process are counted.

4. The online analysis method for the particle size of the material discharged from a mining truck according to claim 1 is characterized by: The workstation in the central control room processes data based on the particle size calculation results and generates a chart showing the proportion of each particle size of the ore and a particle size change curve.