Block rock mineral space relation sensing particle size distribution characteristic measurement method and system

Abstract

The invention relates to the technical field of rock mineral mining, in particular to a blocky rock mineral space relation sensing particle size distribution characteristic measurement method and system, and the method comprises the steps: recognizing the working process of mining equipment, and obtaining equipment operation data, mining surface data and rock sample data; constructing a mining surface recognition model to recognize the mining surface data to obtain mining surface feature data; constructing a rock sample recognition model to recognize the rock sample data to obtain rock sample feature data; setting a sliding time window to collect mining surface feature data, rock sample feature data and equipment operation data to obtain mining surface detection data, rock sample detection data and equipment operation detection data; and constructing an anomaly recognition model to recognize the mining surface detection data, the rock sample detection data and the equipment operation detection data to obtain a mining anomaly coefficient. According to the method, accurate judgment and early warning of the mining abnormity are realized through the mining abnormity coefficient.

Description

Technical Field

[0001] The present invention relates to the technical field of rock and mineral mining, and in particular to a method and system for measuring the spatial relationship perception particle size distribution characteristics of blocky rock and minerals. Background Art

[0002] Based on the needs of traffic planning, energy drilling and mineral mining, people are increasingly dependent on underground resources; for example, digging tunnels underground in cities to build subway systems and improve urban traffic conditions; exploring underground energy materials such as oil and natural gas, and mining various types of ores as raw materials for industrial production; however, due to the existence of various rocks in underground space, such as granite, limestone, shale, sandstone and bedrock, the development of underground resources has been troubled, resulting in low development efficiency and high cost.

[0003] In order to cope with the interference of underground rocks and improve the efficiency of underground resource development, large-scale mining equipment such as tunnel boring machines have been developed. During the working process, the tunnel boring machine uses the high-strength material cutter equipped to cut rocks and soil, and sends the cut materials to the rear through the conveying device; it effectively copes with the interference of rocks and improves the efficiency of underground resource development; but at the same time, the tunnel boring machine also faces safety threats during the working process. When the rock changes, such as changes in rock hardness, cracks and collapses, it is easy to cause damage to the tunnel boring machine.

[0004] Due to the complex design and many components of the tunnel boring machine, its construction and maintenance costs are high. Once damage or failure occurs, it will cause high losses. Therefore, during the operation of the tunnel boring machine, it is necessary to identify and warn of abnormalities in the working process.

[0005] The existing technology for studying rock mining equipment mainly uses the parameters of the mining equipment itself during operation to provide early warning, which can identify dangerous conditions to a certain extent; however, this method fails to combine the characteristics of rock minerals and cannot accurately identify abnormal rock changes, resulting in inaccurate early warning.

[0006] Therefore, a method and system for measuring the particle size distribution characteristics of blocky rock minerals based on spatial relationship perception is proposed. Summary of the invention

[0007] The purpose of the present invention is to provide a method and system for measuring the particle size distribution characteristics of block rock minerals based on spatial relationship perception, to identify the working process of mining equipment, to obtain equipment operation data, mining surface data and rock sample data; to construct a mining surface recognition model to identify the mining surface data, to obtain mining surface characteristic data; to construct a rock sample recognition model to identify the rock sample data, to obtain rock sample characteristic data; to set a sliding time window to collect the mining surface characteristic data, rock sample characteristic data and equipment operation data, to obtain mining surface detection data, rock sample detection data and equipment operation detection data; to construct an anomaly recognition model to identify the mining surface detection data, rock sample detection data and equipment operation detection data, to obtain a mining anomaly coefficient; and to achieve accurate judgment and early warning of mining anomalies through the mining anomaly coefficient.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] The method for measuring the spatial relationship perception particle size distribution characteristics of bulk rock minerals includes:

[0010] S10. Identify the working process of the mining equipment, determine the location of the mining equipment and the mining surface, and collect the mined rock samples; obtain equipment operation data by monitoring the mining equipment, obtain mining surface data by monitoring the mining surface, and obtain rock sample data by the rock samples;

[0011] S20. Constructing a mining face recognition model, identifying the mining face data through the mining face recognition model, and obtaining mining face feature data; the mining face feature data includes geometric feature data, texture feature data, particle feature data and pore feature data;

[0012] S30. Constructing a rock sample identification model; identifying the rock sample data through the rock sample identification model to obtain rock sample characteristic data; the rock sample characteristic data includes component composition data, particle size distribution data, shape distribution data and edge distribution data of the rock sample;

[0013] S40. Set a sliding time window; collect the mining face characteristic data, rock sample characteristic data and equipment operation data through the sliding time window to obtain mining face detection data, rock sample detection data and equipment operation detection data;

[0014] S50. Construct an anomaly recognition model to recognize the mining face detection data, rock sample detection data and equipment operation detection data to obtain a mining anomaly coefficient; and issue an anomaly warning through the mining anomaly coefficient.

[0015] The equipment operation data is obtained by monitoring the mining equipment through sensors, including propulsion force, travel speed, travel direction, power data, tool speed, cutting depth and vibration data;

[0016] The mining surface data is obtained by detecting the mining surface through a high-definition camera, a laser scanning device and an acoustic wave scanning device, and includes mining surface image data, laser scanning data and acoustic wave scanning data;

[0017] The rock sample data is image data of the rock sample, which is obtained by photographing the collected rock sample from multiple directions.

[0018] The mining face recognition model includes a first data input layer, a geometric feature recognition layer, a texture feature recognition layer, a particle feature recognition layer, a pore feature recognition layer and a first data output layer;

[0019] The first data input layer is used to input the mining surface data into the model;

[0020] The geometric feature recognition layer is used to recognize geometric features in the mining surface data to obtain geometric feature data;

[0021] The texture feature recognition layer is used to recognize texture features in the mining surface data to obtain texture feature data;

[0022] The particle feature recognition layer is used to recognize the particle features in the mining surface data to obtain particle feature data;

[0023] The pore feature identification layer is used to identify the pore features in the mining surface data to obtain pore feature data;

[0024] The first data output layer is used to integrate the geometric feature data, texture feature data, particle feature data and pore feature data into mining surface feature data, and output the mining surface feature data.

[0025] The identification process of the particle feature identification layer is as follows:

[0026] Acquire mining surface data, and extract mining surface image data from the mining surface data;

[0027] Identifying regional distribution data of rock particles in the mining face image data, wherein the regional distribution data includes rock particle types, particle regional areas, particle regional shapes, and particle regional distribution;

[0028] The particle characteristic data are obtained based on the mining surface area and the regional distribution data of rock particles.

[0029] The rock sample identification model includes a second data input layer, a component composition identification layer, a particle size distribution identification layer, a shape distribution identification layer, an edge distribution identification layer, and a second data output layer;

[0030] The second data input layer is used to input rock sample data into the model;

[0031] The component composition identification layer is used to identify the component composition in the rock sample data to obtain component composition data;

[0032] The particle size distribution identification layer is used to identify the sample particle size of the rock sample data to obtain particle size distribution data;

[0033] The shape distribution recognition layer is used to recognize the sample shape of the rock sample data to obtain shape distribution data;

[0034] The edge distribution identification layer is used to identify the sample edge of the rock sample data to obtain edge distribution data;

[0035] The second data output layer integrates the component composition data, particle size distribution data, shape distribution data and edge distribution data into rock sample characteristic data, and outputs the rock sample characteristic data.

[0036] The anomaly recognition model is obtained by training a long short-term memory network, and includes a third data input layer, a data feature recognition layer, and a third data output layer;

[0037] The third data input layer inputs mining face detection data, rock sample detection data and equipment operation detection data;

[0038] The data feature recognition layer is used to identify the time variation characteristics of the mining face detection data, rock sample detection data and equipment operation detection data to obtain the mining anomaly coefficient;

[0039] The third data output layer is used to output the mining anomaly coefficient.

[0040] The block rock mineral spatial relationship perception particle size distribution characteristic measurement system includes:

[0041] Data acquisition module: identifies the working process of the mining equipment, determines the location of the mining equipment and the mining surface, and collects the mined rock samples; obtains the equipment operation data by monitoring the mining equipment, obtains the mining surface data by monitoring the mining surface, and detects the rock sample data through the rock samples;

[0042] A mining face recognition module; constructing a mining face recognition model, identifying the mining face data through the mining face recognition model, and obtaining mining face feature data; the mining face feature data includes geometric feature data, texture feature data, particle feature data and pore feature data;

[0043] A rock sample identification module is used to construct a rock sample identification model; the rock sample data is identified by the rock sample identification model to obtain rock sample characteristic data; the rock sample characteristic data includes component composition data, particle size distribution data, shape distribution data and edge distribution data of the rock sample;

[0044] The data processing module sets a sliding time window; collects the mining face characteristic data, rock sample characteristic data and equipment operation data through the sliding time window to obtain mining face detection data, rock sample detection data and equipment operation detection data;

[0045] The anomaly recognition module constructs an anomaly recognition model, and uses the anomaly recognition model to recognize the mining face detection data, rock sample detection data and equipment operation detection data to obtain a mining anomaly coefficient; and uses the mining anomaly coefficient to issue an anomaly warning.

[0046] The equipment operation data is obtained by monitoring the mining equipment through sensors, including propulsion force, travel speed, travel direction, power data, tool speed, cutting depth and vibration data;

[0047] The mining surface data is obtained by detecting the mining surface through a high-definition camera, a laser scanning device and an acoustic wave scanning device, and includes mining surface image data, laser scanning data and acoustic wave scanning data;

[0048] The rock sample data is image data of the rock sample, which is obtained by photographing the collected rock sample from multiple directions.

[0049] The mining face recognition model includes a first data input layer, a geometric feature recognition layer, a texture feature recognition layer, a particle feature recognition layer, a pore feature recognition layer and a first data output layer;

[0050] The first data input layer is used to input the mining surface data into the model;

[0051] The geometric feature recognition layer is used to recognize geometric features in the mining surface data to obtain geometric feature data;

[0052] The texture feature recognition layer is used to recognize texture features in the mining surface data to obtain texture feature data;

[0053] The particle feature recognition layer is used to recognize the particle features in the mining surface data to obtain particle feature data;

[0054] The pore feature identification layer is used to identify the pore features in the mining surface data to obtain pore feature data;

[0055] The first data output layer is used to integrate the geometric feature data, texture feature data, particle feature data and pore feature data into mining surface feature data, and output the mining surface feature data.

[0056] The identification process of the particle feature identification layer is as follows:

[0057] Acquire mining surface data, and extract mining surface image data from the mining surface data;

[0058] Identifying regional distribution data of rock particles in the mining face image data, wherein the regional distribution data includes rock particle types, particle regional areas, particle regional shapes, and particle regional distribution;

[0059] The particle characteristic data are obtained based on the mining surface area and the regional distribution data of rock particles.

[0060] Compared with the prior art, the present invention has the following beneficial effects:

[0061] 1. The present invention constructs a mining face recognition model to recognize mining face data, and recognizes mining face data through dimensions such as geometric features, texture features, particle features and pore features to obtain corresponding geometric feature data, texture feature data, particle feature data and pore feature data; accurately understand the data characteristics of mining face rock.

[0062] 2. The present invention identifies the regional distribution data of rock particles in the mining surface image data to obtain the rock particle type, particle regional area, particle regional shape and particle regional distribution; then obtains the particle characteristic data through the mining surface area and the regional distribution data of rock particles, thereby accurately identifying the rock particle characteristics of the mining surface.

[0063] 3. The present invention constructs a rock sample identification model to identify rock sample data, and identifies the rock sample data from the perspectives of component composition, particle size distribution, shape distribution and edge distribution to obtain corresponding component composition data, particle size distribution data, shape distribution data and edge distribution data; thereby accurately identifying the data characteristics of the rock sample.

[0064] 4. The present invention uses the long short-term memory network as the basis for training to obtain an anomaly recognition model; the time variation characteristics of the mining face detection data, the rock sample detection data and the equipment operation detection data are identified through the anomaly recognition model, and the mining anomaly coefficient is obtained; the mining anomaly coefficient can be used to accurately identify and warn of abnormal mining situations. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a schematic flow chart of the method for measuring the particle size distribution characteristics of block rock minerals by spatial relationship perception of the present invention;

[0066] Figure 2 It is a structural schematic diagram of the mining face identification model of the present invention;

[0067] Figure 3 It is a structural schematic diagram of the rock sample identification model of the present invention;

[0068] Figure 4 It is a structural schematic diagram of the abnormality recognition model of the present invention;

[0069] Figure 5 It is a schematic diagram of the structure of the block rock mineral spatial relationship perception particle size distribution characteristic measurement system of the present invention. DETAILED DESCRIPTION

[0070] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0071] In order to better cope with the changes in the rock environment of a tunnel boring machine during operation, the present invention proposes a method and system for measuring the particle size distribution characteristics of blocky rock minerals through spatial relationship perception.

[0072] Embodiment 1

[0073] The present invention proposes a method for measuring the spatial relationship perception particle size distribution characteristics of block rock minerals. The process of the method is as follows: Figure 1 As shown, including:

[0074] S10. Identify the working process of the mining equipment, determine the positions of the mining equipment and the mining surface, and collect the mined rock samples; obtain equipment operation data by monitoring the mining equipment, obtain mining surface data by monitoring the mining surface, and obtain rock sample data by testing the rock samples.

[0075] The equipment operation data is obtained by monitoring the mining equipment through sensors, including propulsion force, travel speed, travel direction, power data, tool speed, cutting depth and vibration data;

[0076] The propulsion force is the force that pushes the mining equipment forward; the travel speed is the moving speed of the mining equipment; the travel direction is the forward direction of the tunnel boring machine; the power data is the current consumption and power output data of the tunnel boring machine during operation; the tool speed is the rotation speed of the tool; the cutting depth is the depth of the cutting into the rock; the vibration data is the vibration data of the tunnel boring machine in the working state, which is obtained through the sensors set on the tunnel boring machine. Among them, the mining equipment position, mining surface position, mining equipment travel speed and travel direction are obtained by constructing three-dimensional coordinates.

[0077] The mining surface data is obtained by detecting the mining surface through a high-definition camera, a laser scanning device and an acoustic wave scanning device, and includes mining surface image data, laser scanning data and acoustic wave scanning data;

[0078] The rock sample data is image data of the rock sample, which is obtained by photographing the collected rock sample from multiple directions, and the rock sample is obtained by sampling the mined rock fragments.

[0079] The present invention detects mining equipment through sensors to obtain data of the mining equipment in a working state, thereby obtaining equipment operation data; detects the mining surface through a high-definition camera, a laser scanning device and an acoustic wave scanning device to obtain mining surface data; obtains rock sample data by collecting and detecting rock samples; and provides data support for subsequent abnormal identification of mining equipment.

[0080] S20. Construct a mining face recognition model, and use the mining face recognition model to recognize the mining face data to obtain mining face feature data; the mining face feature data includes geometric feature data, texture feature data, particle feature data, and pore feature data.

[0081] The mining face recognition model is constructed based on a deep neural network model, and its structure is as follows: Figure 2 As shown, it includes a first data input layer, a geometric feature recognition layer, a texture feature recognition layer, a particle feature recognition layer, a pore feature recognition layer and a first data output layer;

[0082] The first data input layer is used to input the mining surface data into the model;

[0083] The geometric feature recognition layer is used to recognize geometric features in the mining surface data, wherein the geometric features include the degree of concavity and convexity, inclination angle, convexity, concavity and contour curve of the mining surface, and obtain geometric feature data;

[0084] The texture feature recognition layer is used to recognize texture features in the mining surface data, including texture changes in various areas of the mining surface, and obtain texture feature data;

[0085] The particle feature recognition layer is used to recognize the particle features in the mining face data, including the distribution of rock particles in the mining face, to obtain particle feature data;

[0086] The pore feature identification layer is used to identify the pore features in the mining surface data, including the shape, size and distribution pattern of the pores, and obtain pore feature data;

[0087] The first data output layer is used to integrate the geometric feature data, texture feature data, particle feature data and pore feature data into mining surface feature data, and output the mining surface feature data.

[0088] The present invention constructs a mining face recognition model to recognize mining face data, recognizes mining face data through dimensions such as geometric features, texture features, particle features and pore features, and obtains corresponding geometric feature data, texture feature data, particle feature data and pore feature data; and accurately understands the data features of mining face rock.

[0089] The recognition process of the particle feature recognition layer is as follows:

[0090] Acquire mining surface data, and extract mining surface image data from the mining surface data;

[0091] Identifying regional distribution data of rock particles in the mining face image data, wherein the regional distribution data includes rock particle types, particle regional areas, particle regional shapes, and particle regional distribution;

[0092] The particle characteristic data are obtained based on the mining surface area and the regional distribution data of rock particles.

[0093] The present invention identifies the regional distribution data of rock particles in the mining face image data to obtain the rock particle type, particle regional area, particle regional shape and particle regional distribution; then obtains the particle characteristic data through the mining face area and the regional distribution data of rock particles, thereby accurately identifying the rock particle characteristics of the mining face.

[0094] S30. Construct a rock sample identification model; identify the rock sample data through the rock sample identification model to obtain rock sample characteristic data; the rock sample characteristic data includes component composition data, particle size distribution data, shape distribution data and edge distribution data of the rock sample.

[0095] The rock sample recognition model is constructed based on a deep neural network model, and its structure is as follows: Figure 3 As shown, it includes a second data input layer, a component composition recognition layer, a particle size distribution recognition layer, a shape distribution recognition layer, an edge distribution recognition layer, and a second data output layer;

[0096] The second data input layer is used to input rock sample data into the model;

[0097] The component composition identification layer is used to identify the component composition in the rock sample data to obtain component composition data; it is used to reflect the type and composition of the rock sample;

[0098] The particle size distribution identification layer is used to identify the sample particle size of the rock sample data to obtain particle size distribution data; and is used to analyze the distribution of different rock samples;

[0099] The shape distribution recognition layer is used to recognize the sample shape of the rock sample data to obtain shape distribution data; it is used to reflect the shape distribution and changes of the rock sample;

[0100] The edge distribution identification layer is used to identify the sample edge of the rock sample data to obtain edge distribution data; and is used to reflect the edge characteristics of the rock sample;

[0101] The second data output layer integrates the component composition data, particle size distribution data, shape distribution data and edge distribution data into rock sample characteristic data, and outputs the rock sample characteristic data.

[0102] The present invention constructs a rock sample identification model to identify rock sample data, and identifies the rock sample data from the perspectives of component composition, particle size distribution, shape distribution and edge distribution to obtain corresponding component composition data, particle size distribution data, shape distribution data and edge distribution data; thereby accurately identifying the data characteristics of the rock sample.

[0103] S40. Set a sliding time window; collect the mining face characteristic data, rock sample characteristic data and equipment operation data through the sliding time window to obtain mining face detection data, rock sample detection data and equipment operation detection data.

[0104] S50. Construct an anomaly recognition model to recognize the mining face detection data, rock sample detection data and equipment operation detection data to obtain a mining anomaly coefficient; and issue an anomaly warning through the mining anomaly coefficient.

[0105] The anomaly recognition model is obtained by training the long short-term memory network, and its structure is as follows: Figure 4 As shown, it includes a third data input layer, a data feature recognition layer and a third data output layer;

[0106] The third data input layer inputs mining face detection data, rock sample detection data and equipment operation detection data;

[0107] The data feature recognition layer is used to identify the time variation characteristics of the mining face detection data, rock sample detection data and equipment operation detection data to obtain the mining anomaly coefficient;

[0108] The third data output layer is used to output the mining anomaly coefficient.

[0109] The present invention obtains an anomaly recognition model based on long short-term memory network training; the time variation characteristics of mining face detection data, rock sample detection data and equipment operation detection data are identified through the anomaly recognition model, and the mining anomaly coefficient is obtained; the mining anomaly coefficient can accurately identify and warn of abnormal mining situations.

[0110] By analyzing the data characteristics of the rock on the mining surface and the data characteristics of the mined rock fragments during the operation of the mining equipment, changes in the rock environment can be identified. Combined with the operating data of the mining equipment itself, early warning of abnormalities in the mining equipment can be issued.

[0111] The present invention identifies the working process of mining equipment to obtain equipment operation data, mining face data and rock sample data; constructs a mining face recognition model to identify the mining face data to obtain mining face feature data; constructs a rock sample recognition model to identify the rock sample data to obtain rock sample feature data; sets a sliding time window to collect the mining face feature data, rock sample feature data and equipment operation data to obtain mining face detection data, rock sample detection data and equipment operation detection data; constructs an anomaly recognition model to identify the mining face detection data, rock sample detection data and equipment operation detection data to obtain a mining anomaly coefficient; and realizes accurate judgment and early warning of mining anomalies.

[0112] Embodiment 2

[0113] The present invention proposes a system for measuring the spatial relationship perception particle size distribution characteristics of block rock minerals. The structure of the system is as follows: Figure 5 As shown, it includes a data acquisition module, a mining surface identification module, a rock sample identification module, a data processing module and an anomaly identification module.

[0114] Data acquisition module: identifies the working process of the mining equipment, determines the location of the mining equipment and the mining surface, and collects the mined rock samples; obtains the equipment operation data by monitoring the mining equipment, obtains the mining surface data by monitoring the mining surface, and obtains the rock sample data by detecting the rock samples;

[0115] The equipment operation data is obtained by monitoring the mining equipment through sensors, including propulsion force, travel speed, travel direction, power data, tool speed, cutting depth and vibration data;

[0116] The mining surface data is obtained by detecting the mining surface through a high-definition camera, a laser scanning device and an acoustic wave scanning device, and includes mining surface image data, laser scanning data and acoustic wave scanning data;

[0117] The rock sample data is image data of the rock sample, which is obtained by photographing the collected rock sample from multiple directions.

[0118] A mining face recognition module; constructing a mining face recognition model, identifying the mining face data through the mining face recognition model, and obtaining mining face feature data; the mining face feature data includes geometric feature data, texture feature data, particle feature data and pore feature data;

[0119] The mining face recognition model includes a first data input layer, a geometric feature recognition layer, a texture feature recognition layer, a particle feature recognition layer, a pore feature recognition layer and a first data output layer;

[0120] The first data input layer is used to input the mining surface data into the model;

[0121] The geometric feature recognition layer is used to recognize geometric features in the mining surface data to obtain geometric feature data;

[0122] The texture feature recognition layer is used to recognize texture features in the mining surface data to obtain texture feature data;

[0123] The particle feature recognition layer is used to recognize the particle features in the mining surface data to obtain particle feature data;

[0124] The pore feature identification layer is used to identify the pore features in the mining surface data to obtain pore feature data;

[0125] The first data output layer is used to integrate the geometric feature data, texture feature data, particle feature data and pore feature data into mining surface feature data, and output the mining surface feature data.

[0126] The recognition process of the particle feature recognition layer is as follows:

[0127] Acquire mining surface data, and extract mining surface image data from the mining surface data;

[0128] Identifying regional distribution data of rock particles in the mining face image data, wherein the regional distribution data includes rock particle types, particle regional areas, particle regional shapes, and particle regional distribution;

[0129] The particle characteristic data are obtained based on the mining surface area and the regional distribution data of rock particles.

[0130] A rock sample identification module is used to construct a rock sample identification model; the rock sample data is identified by the rock sample identification model to obtain rock sample characteristic data; the rock sample characteristic data includes component composition data, particle size distribution data, shape distribution data and edge distribution data of the rock sample;

[0131] The rock sample identification model includes a second data input layer, a component composition identification layer, a particle size distribution identification layer, a shape distribution identification layer, an edge distribution identification layer, and a second data output layer;

[0132] The second data input layer is used to input rock sample data into the model;

[0133] The component composition identification layer is used to identify the component composition in the rock sample data to obtain component composition data;

[0134] The particle size distribution identification layer is used to identify the sample particle size of the rock sample data to obtain particle size distribution data;

[0135] The shape distribution recognition layer is used to recognize the sample shape of the rock sample data to obtain shape distribution data;

[0136] The edge distribution identification layer is used to identify the sample edge of the rock sample data to obtain edge distribution data;

[0137] The second data output layer integrates the component composition data, particle size distribution data, shape distribution data and edge distribution data into rock sample characteristic data, and outputs the rock sample characteristic data.

[0138] The data processing module sets a sliding time window; collects the mining face characteristic data, rock sample characteristic data and equipment operation data through the sliding time window to obtain mining face detection data, rock sample detection data and equipment operation detection data;

[0139] The anomaly recognition module constructs an anomaly recognition model, and uses the anomaly recognition model to recognize the mining face detection data, rock sample detection data and equipment operation detection data to obtain a mining anomaly coefficient; and uses the mining anomaly coefficient to issue an anomaly warning.

[0140] The anomaly recognition model is obtained by training a long short-term memory network, and includes a third data input layer, a data feature recognition layer, and a third data output layer;

[0141] The third data input layer inputs mining face detection data, rock sample detection data and equipment operation detection data;

[0142] The data feature recognition layer is used to identify the time variation characteristics of the mining face detection data, rock sample detection data and equipment operation detection data to obtain the mining anomaly coefficient;

[0143] The third data output layer is used to output the mining anomaly coefficient.

[0144] The present invention identifies the working process of mining equipment to obtain equipment operation data, mining surface data and rock sample data; constructs a mining surface recognition model to identify the mining surface data to obtain mining surface feature data; constructs a rock sample recognition model to identify the rock sample data to obtain rock sample feature data; sets a sliding time window to collect the mining surface feature data, rock sample feature data and equipment operation data to obtain mining surface detection data, rock sample detection data and equipment operation detection data; constructs an anomaly recognition model to identify the mining surface detection data, rock sample detection data and equipment operation detection data to obtain a mining anomaly coefficient; and realizes accurate judgment and early warning of mining anomalies through the mining anomaly coefficient.

[0145] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for measuring the spatial relationship perception particle size distribution characteristics of blocky rock minerals, characterized in that: include: S10. Identify the working process of the mining equipment, determine the location of the mining equipment and the mining surface, and collect the mined rock samples; The equipment operation data is obtained by monitoring the mining equipment, the mining surface data is obtained by monitoring the mining surface, and the rock sample data is obtained by the rock sample; S20. Constructing a mining face recognition model, identifying the mining face data through the mining face recognition model, and obtaining mining face feature data; the mining face feature data includes geometric feature data, texture feature data, particle feature data and pore feature data; S30. Constructing a rock sample identification model; identifying the rock sample data through the rock sample identification model to obtain rock sample characteristic data; the rock sample characteristic data includes component composition data, particle size distribution data, shape distribution data and edge distribution data of the rock sample; S40. Set a sliding time window; collect the mining face characteristic data, rock sample characteristic data and equipment operation data through the sliding time window to obtain mining face detection data, rock sample detection data and equipment operation detection data; S50. Construct an anomaly recognition model to recognize the mining face detection data, rock sample detection data and equipment operation detection data to obtain a mining anomaly coefficient; and issue an anomaly warning through the mining anomaly coefficient.

2. The method for measuring the spatial relationship perception particle size distribution characteristics of blocky rock minerals according to claim 1 is characterized by: The equipment operation data is obtained by monitoring the mining equipment through sensors, including propulsion force, travel speed, travel direction, power data, tool speed, cutting depth and vibration data; The mining surface data is obtained by detecting the mining surface through a high-definition camera, a laser scanning device and an acoustic wave scanning device, and includes mining surface image data, laser scanning data and acoustic wave scanning data; The rock sample data is image data of the rock sample, which is obtained by photographing the collected rock sample from multiple directions.

3. The method for measuring the spatial relationship perception particle size distribution characteristics of blocky rock minerals according to claim 1, characterized in that: The mining face recognition model includes a first data input layer, a geometric feature recognition layer, a texture feature recognition layer, a particle feature recognition layer, a pore feature recognition layer and a first data output layer; The first data input layer is used to input the mining surface data into the model; The geometric feature recognition layer is used to recognize geometric features in the mining surface data to obtain geometric feature data; The texture feature recognition layer is used to recognize texture features in the mining surface data to obtain texture feature data; The particle feature recognition layer is used to recognize the particle features in the mining surface data to obtain particle feature data; The pore feature identification layer is used to identify the pore features in the mining surface data to obtain pore feature data; The first data output layer is used to integrate the geometric feature data, texture feature data, particle feature data and pore feature data into mining surface feature data, and output the mining surface feature data.

4. The method for measuring the spatial relationship perception particle size distribution characteristics of blocky rock minerals according to claim 3 is characterized by: The identification process of the particle feature identification layer is as follows: Acquire mining surface data, and extract mining surface image data from the mining surface data; Identifying regional distribution data of rock particles in the mining face image data, wherein the regional distribution data includes rock particle types, particle regional areas, particle regional shapes, and particle regional distribution; The particle characteristic data are obtained based on the mining surface area and the regional distribution data of rock particles.

5. The method for measuring the spatial relationship perception particle size distribution characteristics of blocky rock minerals according to claim 1, characterized in that: The rock sample identification model includes a second data input layer, a component composition identification layer, a particle size distribution identification layer, a shape distribution identification layer, an edge distribution identification layer, and a second data output layer; The second data input layer is used to input rock sample data into the model; The component composition identification layer is used to identify the component composition in the rock sample data to obtain component composition data; The particle size distribution identification layer is used to identify the sample particle size of the rock sample data to obtain particle size distribution data; The shape distribution recognition layer is used to recognize the sample shape of the rock sample data to obtain shape distribution data; The edge distribution identification layer is used to identify the sample edges of the rock sample data to obtain edge distribution data; The second data output layer integrates the component composition data, particle size distribution data, shape distribution data and edge distribution data into rock sample characteristic data, and outputs the rock sample characteristic data.

6. The method for measuring the spatial relationship perception particle size distribution characteristics of blocky rock minerals according to claim 1, characterized in that: The anomaly recognition model is obtained by training a long short-term memory network, and includes a third data input layer, a data feature recognition layer, and a third data output layer; The third data input layer inputs mining face detection data, rock sample detection data and equipment operation detection data; The data feature recognition layer is used to identify the time variation characteristics of the mining face detection data, rock sample detection data and equipment operation detection data to obtain the mining anomaly coefficient; The third data output layer is used to output the mining anomaly coefficient.

7. Block rock mineral spatial relationship perception particle size distribution characteristic measurement system, characterized by: include: Data acquisition module: identifies the working process of the mining equipment, determines the location of the mining equipment and the mining surface, and collects the mined rock samples; The equipment operation data is obtained by monitoring the mining equipment, the mining surface data is obtained by monitoring the mining surface, and the rock sample data is obtained by the rock sample; A mining face recognition module; constructing a mining face recognition model, identifying the mining face data through the mining face recognition model, and obtaining mining face feature data; the mining face feature data includes geometric feature data, texture feature data, particle feature data and pore feature data; A rock sample identification module is used to construct a rock sample identification model; the rock sample data is identified by the rock sample identification model to obtain rock sample characteristic data; the rock sample characteristic data includes component composition data, particle size distribution data, shape distribution data and edge distribution data of the rock sample; The data processing module sets a sliding time window; collects the mining face characteristic data, rock sample characteristic data and equipment operation data through the sliding time window to obtain mining face detection data, rock sample detection data and equipment operation detection data; The anomaly recognition module constructs an anomaly recognition model, and uses the anomaly recognition model to recognize the mining face detection data, rock sample detection data and equipment operation detection data to obtain a mining anomaly coefficient; and uses the mining anomaly coefficient to issue an anomaly warning.

8. The block rock mineral spatial relationship perception particle size distribution characteristic measurement system according to claim 7 is characterized by: The equipment operation data is obtained by monitoring the mining equipment through sensors, including propulsion force, travel speed, travel direction, power data, tool speed, cutting depth and vibration data; The mining surface data is obtained by detecting the mining surface through a high-definition camera, a laser scanning device and an acoustic wave scanning device, and includes mining surface image data, laser scanning data and acoustic wave scanning data; The rock sample data is image data of the rock sample, which is obtained by photographing the collected rock sample from multiple directions.

9. The block rock mineral spatial relationship perception particle size distribution characteristic measurement system according to claim 7, characterized in that: The mining face recognition model includes a first data input layer, a geometric feature recognition layer, a texture feature recognition layer, a particle feature recognition layer, a pore feature recognition layer and a first data output layer; The first data input layer is used to input the mining surface data into the model; The geometric feature recognition layer is used to recognize geometric features in the mining surface data to obtain geometric feature data; The texture feature recognition layer is used to recognize texture features in the mining surface data to obtain texture feature data; The particle feature recognition layer is used to recognize the particle features in the mining surface data to obtain particle feature data; The pore feature identification layer is used to identify the pore features in the mining surface data to obtain pore feature data; The first data output layer is used to integrate the geometric feature data, texture feature data, particle feature data and pore feature data into mining surface feature data, and output the mining surface feature data.

10. The block rock mineral spatial relationship perception particle size distribution characteristic measurement system according to claim 9, characterized in that: The identification process of the particle feature identification layer is as follows: Acquire mining surface data, and extract mining surface image data from the mining surface data; Identifying regional distribution data of rock particles in the mining face image data, wherein the regional distribution data includes rock particle types, particle regional areas, particle regional shapes, and particle regional distribution; The particle characteristic data are obtained based on the mining surface area and the regional distribution data of rock particles.

Images