Mine fully-mechanized face six-machine equipment total monitoring and video linkage analysis system

Abstract

The invention discloses a mine fully-mechanized face six-machine equipment total monitoring and video linkage analysis system, and belongs to the technical field of coal mine fully-mechanized face equipment monitoring. For six-machine equipment, a unified coding specification and an OPC interface + message queue technology are adopted to realize full acquisition and standardized processing of equipment operation data; performing space-time alignment on the equipment state data and the video stream through GIS space coordinate mapping to form a linkage verification mechanism; constructing an idling recognition model based on the load curve and the yield, judging invalid operation and pushing optimization suggestions; and invalid data uploading is reduced through edge preprocessing, and the pressure of the center end is reduced. The problems that traditional monitoring data are heterogeneous, space correlation is weak, efficiency analysis is insufficient and transmission pressure is large are solved, troubleshooting time is shortened, idling energy consumption is reduced, and equipment management efficiency and safety production management and control precision are improved.

Description

Technical Field

[0001] This application relates to the field of coal mine fully mechanized mining equipment monitoring and spatial information fusion technology, specifically a full-scale monitoring and spatial video linkage analysis system for six types of equipment in a fully mechanized mining face. Background Technology

[0002] In the production process of fully mechanized coal mining faces, the "six machines" generally include: coal mining machine, front scraper conveyor, rear scraper conveyor, transfer conveyor, crusher, and belt conveyor. As core production equipment, the monitoring of these machines still faces four key issues that urgently need to be addressed: First, data heterogeneity makes it difficult to unify. Due to the lack of unified standards in coding rules and data formats for the "six machines" equipment from different manufacturers, data cannot be aggregated and analyzed across devices. For example, the current data of the Shanghai Chuangli coal mining machine is coded as "CM-001" and that of the German DBT scraper conveyor is coded as "SG-01", and the data of the two cannot be directly correlated.

[0003] Secondly, there is a disconnect between space and video. Existing monitoring systems can only display equipment data (such as abnormal current) or video footage separately. They cannot locate specific spatial positions through equipment data and simultaneously retrieve corresponding area videos. During troubleshooting, manual matching of "data-location-video" is required, which is time-consuming.

[0004] Third, the efficiency of idling identification is low. Traditional identification methods only judge the effectiveness of operation based on the on / off status of the equipment, and cannot accurately distinguish between the two states of "effective operation (coal cutting with load)" and "idling (operation without coal)". This not only causes waste of equipment energy, but also fails to provide targeted suggestions for subsequent efficiency optimization.

[0005] Fourth, the data transmission pressure is relatively high. Currently, all device data is directly uploaded to the central end for processing without preprocessing at the edge. Among them, the proportion of invalid data such as duplicate data during normal device operation is high, which leads to a significant increase in the processing pressure at the central end and an increase in data processing delay.

[0006] Although existing technologies (such as some traditional technologies that only focus on fault monitoring of a single coal mining machine and do not cover the integration of all data from the "six machines"; and some traditional technologies that involve industrial video surveillance but do not combine GIS spatial positioning to achieve "data-location-video" linkage) have not solved the problems of unified coding, GIS-video linkage, intelligent identification of idle time and edge preprocessing integration in the scenario of full monitoring of the "six machines", they are difficult to meet the actual needs of efficient management and performance optimization of fully mechanized mining faces. Summary of the Invention

[0007] In view of the above problems, this application proposes a full-scale monitoring and spatial video linkage analysis system for six types of equipment in a fully mechanized mining face to overcome the shortcomings of the prior art.

[0008] In a first aspect, embodiments of this application provide a full-scale monitoring and video linkage analysis system for six-machine equipment in a fully mechanized mining face, including: a data acquisition module, a data processing module, and a linkage analysis module; The data acquisition module is used to collect full data from the six machines in the fully mechanized mining face. It uses a unified coding standard and OPC interface plus message queue technology to collect equipment operation data. The six machines include: coal mining machine, front scraper conveyor, rear scraper conveyor, transfer conveyor, crusher, and belt conveyor. The data processing module is used to standardize the collected operating data of the six machines and to align the real-time status data of the equipment with the video stream of the working face roadway using GIS spatial coordinate mapping technology. The linkage analysis module is used to dynamically mark the location of abnormal equipment on the mine map and automatically retrieve the corresponding video footage to form a linkage verification mechanism. At the same time, it constructs an idling identification model based on the equipment load curve and production data in the operation data, judges the invalid operation state of the equipment, and pushes efficiency optimization suggestions. The linkage verification mechanism refers to the mechanism of displaying an abnormal alarm based on its real-time operation data, spatial location, and related video when any of the six machines malfunctions.

[0009] Optionally, the unified coding standard is used to achieve unified identification of equipment data from different manufacturers, and it includes the unique identifier of each of the six machines and the coding rules for measurement points; The unique identifier is encoded in the format of device type code + region code + serial number, totaling 18 bits; The measurement point coding rule is: device code + measurement point type code + measurement point serial number.

[0010] Optionally, the message queue of the data acquisition module uses a Kafka cluster; The Kafka cluster supports high-concurrency writing of various types of data and ensures data transmission integrity through a data verification mechanism, including CRC32 checksum.

[0011] Optionally, the standardization processing of the data processing module includes: data format conversion, unit unification, and outlier filtering; The data format is converted to convert proprietary protocol data from different manufacturers into JSON format; The units mentioned are uniformly converted from tons per hour to kilograms per minute; The outlier filtering is based on the 3σ principle to remove data that exceeds a reasonable range.

[0012] Optionally, the GIS spatial coordinate mapping technology is as follows: real-time coordinates are obtained through the UWB positioning module built into the device, with a positioning accuracy of ≤0.5 meters; The GIS spatial coordinate mapping technology is associated with the preset position information of the video surveillance cameras to automatically match the positions of each of the six devices with the video footage.

[0013] Optionally, when the linkage verification mechanism detects an abnormal device current, it completes the following operations within 10 seconds: The location of abnormal equipment is highlighted on the GIS mining map, and more than three video cameras in the area are automatically called. Video clips of the abnormal moment are captured and displayed synchronously with the data curve, which is the curve corresponding to the operating data of the six abnormal machines.

[0014] Optionally, the idling identification model determines the ineffective operating state of the equipment by calculating the idling coefficient of each of the six machines. The idling coefficient is the ratio of operating time to output and is related to the load correction coefficient. When the idling coefficient exceeds a preset threshold, the six machines exceeding the preset threshold are determined to be in an invalid operating state.

[0015] Optionally, the performance optimization suggestions include: equipment parameter adjustment, start-up and shutdown sequence optimization, and maintenance cycle reminders; The equipment parameter adjustments include, but are not limited to: reducing the crusher frequency from 50Hz to 45Hz; The optimization of start-stop timing includes, but is not limited to: the scraper conveyor starts 30 seconds later than the coal mining machine; The maintenance cycle reminder includes, but is not limited to: it is recommended to change the lubricating oil of the transfer machine every 15 days.

[0016] Optionally, the data processing module supports the deployment of edge computing nodes; The edge computing nodes are deployed in control cabinets near the work surface to complete data preprocessing. The data preprocessing reduces invalid data uploaded to the central end. The data preprocessing includes: real-time filtering and local anomaly detection.

[0017] Optionally, it may also include: a device health assessment module; The equipment health assessment module calculates the health index of each six-machine device based on the cumulative running time, number of failures, and load fluctuation range parameters of each six-machine device using the analytic hierarchy process. The equipment health index ranges from 0 to 100 points. When the equipment health index of any six-machine device is lower than 60 points, a maintenance warning is automatically triggered.

[0018] Optionally, the dynamic annotation of the mining map in the linkage analysis module supports multi-layer display, and the dynamic annotation of the mining map can trace back the status and video footage of each of the six machines at any time through the time axis. The multi-layer includes: equipment location layer, data anomaly layer, and video overlay layer.

[0019] Optionally, the preset threshold used in the idling detection model is a dynamically adjustable judgment threshold; the method by which the idling detection model obtains the judgment threshold includes: The coal seam hardness, moisture content, and dip angle are collected using a data acquisition unit deployed on the working face support. The data acquisition unit includes a coal seam hardness sensor and a moisture content sensor. A weighted Naive Bayes fusion algorithm is used to convert the coal seam hardness, moisture content, and dip angle into load correction coefficients. Each load correction coefficient is equal to the sum of the coal seam hardness correction value, moisture content correction value, and dip angle correction value, divided by 3. Specifically: when the coal seam hardness is less than or equal to 2 MPa, the hardness correction value is 1.0; when it is between 2 and 4 MPa, the hardness correction value is 0.8; and when it is greater than 4 MPa, the hardness correction value is 0.6. Similarly, when the moisture content is less than or equal to 8%, the moisture content correction value is 1.0; when it is between 8% and 15%, the moisture content correction value is 0.9; and when it is greater than 15%, the moisture content correction value is 0.7. Finally, when the dip angle is less than or equal to 15°, the dip angle correction value is 1.0; when it is between 15° and 30°, the dip angle correction value is 0.85; and when it is greater than 30°, the dip angle correction value is 0.75. The judgment threshold is dynamically adjusted according to the equipment type and load correction coefficient. Specifically, for a coal mining machine, the judgment threshold is 0.3 when the load correction coefficient is equal to 1.0, and is increased to 0.45 when the correction coefficient is equal to 0.6. For a scraper conveyor, the judgment threshold is 0.28 when the load correction coefficient is equal to 1.0, and is increased to 0.38 when the correction coefficient is equal to 0.75. The idling coefficient is equal to the ratio of the equipment's running time to its output. When the idling coefficient is multiplied by the load correction coefficient, and the idling coefficient exceeds the corresponding judgment threshold, the six machines that exceed the judgment threshold are determined to be in an invalid operating state.

[0020] Optionally, when the linkage verification mechanism detects an abnormal equipment current, it uses a spatiotemporal tracing module to highlight the location of the abnormal equipment on the GIS mining map, automatically call more than 3 video cameras in the area, capture video clips at the time of the abnormality, and display them synchronously with the data curve. The spatiotemporal tracing module includes: an equipment correlation modeling unit, a spatiotemporal causal calculation unit, and a tracing result visualization unit. The equipment association modeling unit constructs a spatiotemporal association matrix based on the physical connection relationship and historical fault data of the six equipments, and marks the influence delay and influence intensity between each equipment. The spatiotemporal causal calculation unit uses a sliding time window + causal entropy algorithm to extract the running data from 60 seconds before the abnormality to 30 seconds after the abnormality when any device malfunctions, calculate the causal entropy value of each device abnormality event, and determine the device with a causal entropy value less than 0.3 as the root cause device. The source tracing result visualization unit marks the root cause device with red flashing, the primary influencing device with orange solid light, and the secondary influencing device with yellow solid light on the GIS mining map. It also generates an anomaly propagation timeline to show the time of anomaly occurrence, impact delay, and data change trend of each device, and synchronously displays video clips and data curves at the time of anomalies.

[0021] Optionally, the edge computing node includes: an edge-end intelligent preprocessing system based on time-series prediction; The edge-end intelligent preprocessing system includes: a time-series data prediction unit, a predictive filtering unit, and a local early warning push unit; The time-series data prediction unit uses an LSTM + attention mechanism to build a lightweight prediction model. Based on the operating data of each device over the past 5 minutes, it predicts the data flow trend for the next 1 minute and outputs the predicted value and confidence interval. The lightweight prediction model uses int8 precision quantization compression to adapt to edge industrial control computers, and the data processing time for a single device is less than or equal to 50ms. The predictive filtering unit follows the following rules: if the predicted value deviates from the current data by less than 5% and the predicted trend is returning to normal, the locally stored data is not uploaded to the central end; if the predicted value exceeds the normal range, the predicted abnormal data and the current data are packaged and uploaded to the central end. The local early warning push unit triggers a local audible and visual early warning when an anomaly is predicted.

[0022] Optionally, the equipment health assessment module is a dynamic health assessment module with two dimensions: equipment type and operating condition. It includes: an equipment type weight library, an operating condition level classification unit, a dynamic weight calculation unit, and a health index correction unit. The equipment type weight library sets differentiated basic weights for the six types of equipment, including: coal mining machine: 25% for running time, 35% for failure frequency, 20% for load fluctuation, and 20% for cutter tooth wear; crusher: 20% for running time, 40% for failure frequency, 15% for load fluctuation, and 25% for bearing temperature; belt conveyor: 30% for running time, 35% for failure frequency, 25% for load fluctuation, and 10% for belt tension. The operating condition classification unit divides each device into three levels based on its real-time load rate: low load (load rate < 60%), medium load (60% ≤ load rate ≤ 90%), and high load (load rate > 90%). The adjustment rules for the dynamic weight calculation unit are as follows: under high load conditions, the weight of load fluctuation is increased by 20% and the weight of failure frequency is increased by 10%; under low load conditions, the weight of running time is increased by 15% and the weight of newly added core parameters is increased by 10%. The core parameters include: the degree of wear of the cutting teeth, the bearing temperature, and the belt tension. The health index correction unit introduces a working condition influence coefficient, setting it to 1.0 for low load, 1.05 for medium load, and 1.1 for high load. The final health index = basic health index × working condition influence coefficient, and the basic health index is calculated based on the adjusted weights.

[0023] Compared with the prior art, the positive effects of this application are: The comprehensive monitoring and spatial video linkage analysis system for six types of equipment in fully mechanized mining faces proposed in this application can effectively solve several key problems in the monitoring of these six types of equipment, providing strong technical support for efficient management and performance optimization of fully mechanized mining faces. This includes addressing data heterogeneity issues, achieving spatial-video linkage, accurately identifying idling status, and reducing data transmission pressure: Firstly, it solves the data heterogeneity problem by using a unified coding standard to achieve unified identification of data from six types of equipment from different manufacturers. This improves data integration efficiency, shortens the time required for cross-equipment data analysis, and solves the problem of difficulty in summarizing and analyzing data from different manufacturers across equipment.

[0024] Secondly, it enables spatial-video linkage. By leveraging GIS mapping and automatic video matching functions, it can shorten the troubleshooting time, improve the accuracy of fault location, and enhance the efficiency and precision of troubleshooting.

[0025] Third, the accurate identification of idling status can distinguish between effective operation and idling status of equipment through the idling identification model, reduce the energy consumption of equipment idling, save electricity costs for each mine, and reduce equipment energy waste.

[0026] Fourth, reducing data transmission pressure can reduce invalid data uploads through edge preprocessing, shorten the processing delay at the central end, and at the same time, equipment health assessment can provide early warning of faults, reduce unplanned downtime, improve data processing efficiency, and ensure production continuity.

[0027] This application proposes a comprehensive monitoring and spatial-video linkage analysis system for the "six machines" (machines, machinery, and equipment) in a fully mechanized mining face. The system aims to provide a full-process, intelligent solution for monitoring the "six machines" in a fully mechanized mining face: by constructing a unified data coding standard, it breaks down data barriers between different manufacturers' equipment, enabling efficient aggregation and correlation analysis of cross-equipment data; relying on the deep linkage between GIS spatial positioning technology and video monitoring systems, it achieves automatic matching of abnormal equipment data with corresponding spatial locations and on-site videos, significantly shortening fault diagnosis time and improving emergency response efficiency; it introduces an intelligent idling identification algorithm to accurately distinguish between effective equipment operation and idling states, providing data support for energy consumption management and simultaneously pushing targeted efficiency optimization suggestions; it adds an edge-end data preprocessing module to filter invalid and duplicate data, reducing the processing pressure on the central end and lowering data transmission latency, ultimately helping the fully mechanized mining face achieve efficient management, energy consumption optimization, and safe and stable operation of the "six machines," meeting the high-quality equipment monitoring requirements of modern coal mine production.

[0028] The system proposed in this application is a system for full-data monitoring, spatial-video linkage verification, and idling identification of the "six machines" (machines, machinery, and equipment) in fully mechanized mining faces. It is particularly suitable for scenarios involving the full lifecycle monitoring, energy consumption optimization, fault early warning, and safety production control of the "six machines" in fully mechanized mining faces. By integrating standardized data processing, GIS spatial positioning, intelligent algorithm recognition, and edge computing technology, it forms a monitoring system covering the entire operating cycle of the "six machines." It not only addresses the technical pain points of each individual problem but also achieves deep integration of monitoring data, spatial location, and video footage through multi-module collaboration. This provides integrated technical support for the management of fully mechanized mining face equipment, from data acquisition and analysis to decision support, filling the current gap in the field of full-data monitoring of the "six machines." Attached Figure Description

[0029] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a modular schematic diagram of a full-scale monitoring and spatial video linkage analysis system for six types of equipment in a fully mechanized mining face, as proposed in an embodiment of this application. Figure 2 This is an architecture diagram of an exemplary mine fully mechanized mining face six-machine equipment full-scale monitoring and spatial video linkage analysis system in the embodiments of this application; Figure 3 This is an overview diagram of the implementation process of the full-scale monitoring and spatial video linkage analysis system for six types of equipment in a fully mechanized mining face, as illustrated in the embodiments of this application. Figure 4This is a schematic diagram of the system information flow of the full-scale monitoring and spatial video linkage analysis system for six types of equipment in a fully mechanized mining face, as illustrated in the embodiments of this application. Detailed Implementation

[0030] The embodiments of this application will now be described in detail. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain this application, and should not be construed as limiting this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0031] This application proposes a comprehensive monitoring and spatial video linkage analysis system for six types of equipment in a fully mechanized mining face, referring to... Figure 1 The modular diagram shown includes: a data acquisition module, a data processing module, and a linkage analysis module.

[0032] The data acquisition module is used to collect full data from the six machines in the fully mechanized mining face. It adopts a unified coding standard and OPC interface plus message queue technology to collect equipment operation data. The six machines generally include: coal mining machine, front scraper conveyor, rear scraper conveyor, transfer conveyor, crusher, and belt conveyor.

[0033] The data processing module is used to standardize the collected operating data of the six machines and to align the real-time status data of the equipment with the video stream of the working face roadway using GIS spatial coordinate mapping technology.

[0034] The linkage analysis module is used to dynamically mark the location of abnormal equipment on the mine map and automatically retrieve the corresponding video footage to form a linkage verification mechanism. At the same time, it builds an idling identification model based on the equipment load curve and production data in the operation data, judges the invalid operation status of the equipment, and pushes efficiency optimization suggestions. The linkage verification mechanism refers to the mechanism that displays an abnormal alarm based on its real-time operation data, spatial location, and related video when any of the six machines malfunctions.

[0035] In one embodiment of this application, combined with Figure 2 The diagram shows an exemplary architecture of a full-scale monitoring and spatial video linkage analysis system for six types of equipment in a fully mechanized mining face. The specific structure and functions of each part are described below: The data acquisition module implements the formulation of unified coding standards and the application of data acquisition technology. The unified coding standards are to formulate 18-bit unique codes for six types of equipment (format: equipment type code + area code + serial number, such as the coal mining machine code "CM-03-140211006496") and measurement point coding rules (such as "CM-03-140211006496-C01-001" represents the No. 1 current measurement point of a specific coal mining machine).

[0036] The data acquisition technology uses OPC interface plus Kafka message queue technology (Kafka cluster supports high-concurrency writing of device data, with CRC32 data verification mechanism and verification failure rate <0.1%) to collect operating data of six machines (such as coal mining machine current, scraper conveyor speed, etc.), GIS spatial data (equipment UWB positioning coordinates), and industrial video stream data (roadway belt, working face support camera images), to achieve unified identification of equipment data from different manufacturers, efficient collection of full data and guarantee of transmission integrity.

[0037] In one embodiment of this application, the data processing module performs standardization and edge preprocessing on the collected data from six machines to reduce the transmission and processing pressure at the central end. This includes two main functions: standardization and edge preprocessing. Standardization involves data format conversion, converting data from different manufacturers' proprietary protocols into JSON format (e.g., converting binary data from a German DBT scraper conveyor to a rotation speed of 1.2 m / s); unit standardization (e.g., converting 300 tons per hour to 5000 kilograms per minute); and outlier filtering, removing unreasonable data based on the 3σ principle (e.g., belt conveyor rotation speeds > 3 m / s or < 0.5 m / s).

[0038] The data processing module supports edge computing node deployment. Edge computing nodes are deployed in the control cabinet near the work surface to perform data preprocessing. This preprocessing reduces invalid data uploaded to the central end. Data preprocessing includes real-time filtering and local anomaly detection. Ideally, real-time filtering (removing noise from current data) can be performed using Kalman filtering, and local anomaly detection based on current exceeding 110% of the rated value (triggering an initial warning) can be implemented. This achieves data format standardization, invalid data filtering, and edge interception, reducing the transmission pressure on the central end.

[0039] A preferred approach is that the edge computing nodes include an edge-end intelligent preprocessing system based on time-series prediction.

[0040] The edge-end intelligent preprocessing system includes: a time-series data prediction unit, a predictive filtering unit, and a local early warning push unit. The time-series data prediction unit uses an LSTM + attention mechanism to build a lightweight prediction model. Based on the operating data of each device over the past 5 minutes, it predicts the data flow trend for the next 1 minute and outputs the predicted value and confidence interval. The lightweight prediction model uses int8 precision quantization compression to adapt to edge industrial control computers, and the data processing time for a single device is less than or equal to 50ms.

[0041] The predictive filtering unit follows these rules: if the predicted value deviates from the current data by less than 5% and the predicted trend is returning to normal, the locally stored data is not uploaded to the central server; if the predicted value exceeds the normal range, the abnormal predicted data plus the current data are packaged and uploaded to the central server. The local early warning push unit triggers a local audible and visual warning when an abnormal prediction is made.

[0042] In one embodiment of this application, the linkage analysis module is used to realize the linkage analysis and verification of data, spatial position and video images of six machines, solve the problem of spatial-video disconnection and accurately identify idle state. It includes: GIS spatial coordinate mapping, "data-space-video" linkage verification and idle identification model. The GIS spatial coordinate mapping is: real-time coordinates are obtained through the built-in UWB positioning module of the equipment, with a positioning accuracy of ≤0.5 meters, and associated with the preset position information of the video camera (e.g., the preset position coordinates of the camera of the 30# support of the fully mechanized mining face X: 517798.265, Y: 4410023.111) to automatically match the position of each of the six machines with the video image.

[0043] The "data-space-video" linkage verification mechanism completes the following within 10 seconds when equipment malfunctions: highlight the abnormal location in red on the GIS mining map, automatically call three regional video cameras, and capture video clips of the abnormal moment and display them synchronously with data curves (the data curves are the curves corresponding to the operating data of the six abnormal machines). For example, when the current of the front scraper conveyor exceeds the rated value of 216A, the location of the No. 10-15 support of the fully mechanized mining face is marked and the abnormal moment video and the corresponding current curve are displayed synchronously.

[0044] A preferred approach is as follows: When the linkage verification mechanism detects an abnormal equipment current, it can use the spatiotemporal tracing module to highlight the location of the abnormal equipment on the GIS mining map, automatically call up more than 3 video cameras in the area, capture video clips at the time of the abnormality, and display them synchronously with the data curve. The spatiotemporal tracing module includes: an equipment correlation modeling unit, a spatiotemporal causal calculation unit, and a tracing result visualization unit.

[0045] The equipment correlation modeling unit constructs a spatiotemporal correlation matrix based on the physical connection relationship of the six equipment (e.g., coal mining machine → front scraper conveyor → transfer machine → belt conveyor) and historical fault data, and marks the influence delay and influence intensity between each equipment.

[0046] The spatiotemporal causal calculation unit uses a sliding time window + causal entropy algorithm to extract the running data from 60 seconds before the abnormality to 30 seconds after the abnormality when any device malfunctions. It calculates the causal entropy value of each device abnormality event and identifies devices with a causal entropy value less than 0.3 as the root cause device.

[0047] The source tracing results visualization unit marks the root cause device with red flashing, the primary affected device with orange solid light, and the secondary affected device with yellow solid light on the GIS mine map. It also generates an anomaly propagation timeline to show the time of anomaly occurrence, impact delay, and data change trend of each device, and synchronously displays video clips and data curves at the time of anomalies.

[0048] The idling identification model determines whether the equipment is running ineffectively by calculating the idling coefficient between the equipment's running time and output. When the idling coefficient exceeds a preset threshold (which is dynamically adjustable, as explained below), the six machines exceeding the preset threshold are determined to be in an ineffective operating state (for example, when the output is 500 tons after 2 hours of operation, the coefficient is 0.5 and it is determined to be idling). This enables automatic matching of equipment location and video, intuitive verification of fault scenarios, and accurate identification of idling status and performance warning (pushing optimization suggestions such as "adjust the angle of the coal mining machine's cutting teeth").

[0049] In one embodiment of this application, the preset threshold used in the idling detection model is a dynamically adjustable judgment threshold; the method by which the idling detection model obtains this dynamically adjustable judgment threshold includes: First, the coal seam hardness, moisture content, and dip angle are collected using a data acquisition unit deployed on the working face support. The data acquisition unit includes a coal seam hardness sensor and a moisture content sensor.

[0050] A weighted Naive Bayes fusion algorithm is used to convert coal seam hardness, moisture content, and dip angle into load correction coefficients. The load correction coefficient is equal to the sum of the coal seam hardness correction value, moisture content correction value, and dip angle correction value, divided by 3, i.e.: Load correction coefficient = (coal seam hardness correction value + moisture content correction value + dip angle correction value) / 3. Specifically, when the coal seam hardness is less than or equal to 2 MPa, the coal seam hardness correction value is 1.0; for values ​​between 2 and 4 MPa... When the pressure is 'a', the coal seam hardness correction value is 0.8; when it is greater than 4 MPa, the coal seam hardness correction value is 0.6. When the moisture content is less than or equal to 8%, the moisture content correction value is 1.0; when it is between 8% and 15%, the moisture content correction value is 0.9; when it is greater than 15%, the moisture content correction value is 0.7. When the dip angle is less than or equal to 15°, the dip angle correction value is 1.0; when it is between 15° and 30°, the dip angle correction value is 0.85; when it is greater than 30°, the dip angle correction value is 0.75. The judgment threshold is dynamically adjusted based on the equipment type and load correction coefficient. Specifically, for coal mining machines, the judgment threshold is 0.3 when the load correction coefficient is 1.0, and it is increased to 0.45 when the correction coefficient is 0.6. For scraper conveyors, the judgment threshold is 0.28 when the load correction coefficient is 1.0, and it is increased to 0.38 when the correction coefficient is 0.75.

[0051] Finally, the obtained idling coefficient is equal to the ratio of the equipment's running time to its output, and the result of multiplying it by the load correction coefficient is: Idle coefficient = (equipment running time / output) × load correction coefficient; when the idling coefficient exceeds the corresponding judgment threshold, the six machines that exceed the judgment threshold are determined to be in an invalid operating state.

[0052] In one embodiment of this application, the dynamic annotation of the mine map in the linkage analysis module supports multi-layer display, and the dynamic annotation of the mine map can trace back the status and video footage of each of the six machines at any time through the time axis. The multi-layer includes: equipment location layer, data anomaly layer, and video overlay layer.

[0053] In one embodiment of this application, the performance optimization suggestions include: equipment parameter adjustment, start-stop timing optimization, and maintenance cycle reminders; equipment parameter adjustment includes, but is not limited to, reducing the crusher frequency from 50Hz to 45Hz, etc.; start-stop timing optimization includes, but is not limited to, the scraper conveyor starting 30 seconds later than the coal mining machine, etc.; maintenance cycle reminders include, but are not limited to, suggesting that the transfer machine's lubricating oil be changed after 15 days, etc.

[0054] In one embodiment of this application, the full-scale monitoring and spatial video linkage analysis system for six machines in a fully mechanized mining face further includes: an equipment health assessment module. The equipment health assessment module is used to assess the health status of the six machines based on their operating parameters and trigger maintenance warnings. It includes: a health index calculation model and a maintenance warning mechanism. The health index calculation model is based on parameters such as cumulative operating time, number of failures, and load fluctuation amplitude. It uses the analytic hierarchy process (AHP) (which assigns weights and then analyzes based on different weights; the optimal weight allocation is: operating time 30%, number of failures 40%, load fluctuation 30%) to calculate a health index of 0-100 points. For example, if a scraper conveyor has an operating time of 1800 hours with a score of 80, 2 failures with a score of 70, and a load fluctuation of ±10% with a score of 90, the health index = 80 × 0.3 + 70 × 0.4 + 90 × 0.3 = 79 points. The maintenance warning mechanism automatically triggers a warning when the health index is below 60 points, achieving quantitative assessment of equipment health status and precise maintenance prompts, such as "It is recommended to replace the scraper chain in 10 days."

[0055] A preferred approach is to have a dynamic health assessment module that is a dual-dimensional assessment module for equipment type and operating conditions, which includes: an equipment type weight library, an operating condition level classification unit, a dynamic weight calculation unit, and a health index correction unit.

[0056] The equipment type weight library sets differentiated basic weights for the six types of equipment, including: coal mining machine: 25% for running time, 35% for failure frequency, 20% for load fluctuation, and 20% for cutter tooth wear; crusher: 20% for running time, 40% for failure frequency, 15% for load fluctuation, and 25% for bearing temperature; belt conveyor: 30% for running time, 35% for failure frequency, 25% for load fluctuation, and 10% for belt tension.

[0057] The operating condition classification unit is divided into three levels based on the real-time load rate of each device: low load, i.e., load rate < 60%; medium load, i.e., 60% ≤ load rate ≤ 90%; and high load, i.e., load rate > 90%.

[0058] The dynamic weight calculation unit adjustment rules are as follows: under high load conditions, the weight of load fluctuation is increased by 20% and the weight of failure frequency is increased by 10%; under low load conditions, the weight of running time is increased by 15% and the weight of newly added core parameters is increased by 10%. These core parameters include: cutter wear degree, bearing temperature, and belt tension.

[0059] The health index correction unit introduces the operating condition influence coefficient, setting it to 1.0 for low load, 1.05 for medium load, and 1.1 for high load. The final health index = basic health index × operating condition influence coefficient, which is calculated based on the adjusted weights.

[0060] In summary, the comprehensive monitoring and spatial video linkage analysis system for six types of equipment in a fully mechanized mining face integrates standardized data collection, collaborative data processing, spatial-video linkage analysis, intelligent identification of idling, and equipment health assessment. Through a framework of "standardized data collection, collaborative data processing, visualized linkage analysis, and precise efficiency optimization," coupled with unified coding standards, OPC interface + Kafka acquisition technology, edge preprocessing, UWB positioning linkage video, and analytic hierarchy process (AHP) health assessment, the system achieves unified identification of equipment data from different manufacturers, significantly reduced fault diagnosis time, accurate determination of idling status, and quantitative assessment of equipment health status (health index 0-100 points). The main purpose of this system is to provide a full-process intelligent solution for the management of six types of equipment in fully mechanized mining faces, helping to break down data heterogeneity barriers, solve the pain point of spatial-video disconnection, reduce ineffective energy consumption and waste, and reduce the data processing pressure at the central end. At the same time, it reduces equipment downtime through precise maintenance and early warning, ensuring the improvement of production efficiency of fully mechanized mining faces, safe and stable operation of equipment, and the realization of the goal of efficient management and control of modern coal mines.

[0061] In the specific implementation process, in conjunction with reference Figure 3 The overview diagram and Figure 4 The system information flow diagram shown can be summarized into the following six steps: Step 1: System Deployment System deployment is the foundation for realizing full monitoring of six machines and spatial-video linkage analysis. It is necessary to build an operating environment adapted to the fully mechanized mining face environment from both hardware and software aspects to ensure the stable implementation of data acquisition, processing, and linkage analysis functions. Figure 3 Step 1 can be summarized as follows: Hardware: Sensors + Edge Nodes; Software: Acquisition / Processing Programs.

[0062] 1) Hardware deployment Front-end data acquisition hardware is deployed in the longwall mining face of the coal mine: data acquisition sensors for six types of equipment (e.g., coal mining machine current sensor; scraper conveyor speed sensor; crusher temperature sensor), used to collect equipment operating parameters; UWB positioning devices are deployed on multiple supports on the working face, such as supports #10, #20, and #30, to obtain the real-time spatial coordinates of the equipment; 4K industrial video cameras, such as one each on supports #10, #15, and #20 and the belt conveyor head, with a dustproof and waterproof rating of IP67, supporting low-light shooting, and real-time acquisition of video footage from the working face.

[0063] Edge computing nodes (e.g., Advantech 610L industrial PCs with I7-12700 processors, 32GB RAM, 1TB SSDs, and dual network cards) are deployed in the control cabinet near the longwall mining face for local data preprocessing. A GIS system (e.g., Longsoft GIS 3.5, supporting LFM / LFMX format mine map loading) and large-screen display devices (e.g., 55-inch video wall with 3840*2160 resolution) are deployed in the mine dispatch room to display spatial-video linkage results and equipment health status.

[0064] 2) Software deployment (1) Data acquisition module: It is developed using Java language, based on the .NET Framework, supports the OPC UA industrial protocol, and is compatible with the data formats of six-machine equipment from different manufacturers such as Shanghai Chuangli and German DBT, so as to realize full data acquisition.

[0065] (2) Data processing module: integrates Kalman filter algorithm library (for noise filtering of current and speed data), JSON format conversion tool (to convert private protocol data into standard format, i.e., standardization processing), develops edge preprocessing program, supports local anomaly judgment and invalid data filtering, designs spatiotemporal alignment algorithm, and uses GIS spatial coordinate mapping technology to perform spatiotemporal alignment of real-time equipment status data and working face roadway video stream.

[0066] (3) Linkage Analysis Module: Developed based on WebGIS technology, it performs GIS spatial dynamic annotation, video linkage retrieval, idle time identification module, and efficiency optimization suggestion output. It supports dynamic annotation of mine maps, automatic retrieval of video footage, synchronous display of data curves and videos, and is adapted to the mine scheduling large screen display.

[0067] (4) Equipment health assessment module: It is developed using Java language and integrates the Analytic Hierarchy Process (AHP) algorithm package, which can calculate the equipment health index and generate maintenance early warning suggestions.

[0068] Step 2: Data Collection and Standardization Data collection and standardization are key to eliminating the problem of heterogeneous data from six different equipment. By unifying coding and standardizing the collection process, it is ensured that data from equipment from different manufacturers can be aggregated and analyzed. Figure 3 Step 2 can be summarized as follows: data format standardization; sampling frequency calibration; outlier filtering.

[0069] 1) Uniform encoding assignment Each of the six types of equipment in a fully mechanized coal mining face is assigned a unique code. The code format is "equipment type code + area code + serial number" (18 digits in total). For example: the coal mining machine code is "CM—03—140211006496", where CM is the coal mining machine type code, 03 is the coal mine area code, and the last 12 digits are the equipment's factory serial number; the front scraper conveyor code is "SG—03—140211006497"; the rear scraper conveyor code is "SG—03—140211006498"; the crusher code is "PC—03—140211006499"; the transfer conveyor code is "ZZ—03—140211006500"; the belt conveyor code is "PD—03—140211006501", etc.

[0070] Establish measurement point coding rules, with the format "Equipment Code - Measurement Point Type Code - Measurement Point Serial Number": For example, the code for the No. 1 current measurement point of the coal mining machine is "CM-03-140211006496-C01-001", where C01 is the current measurement point type code; the code for the No. 1 speed measurement point of the scraper conveyor is "SG-03-140211006497-S01-001", where S01 is the speed measurement point type code, to achieve unique identification of measurement point data.

[0071] 2) Data Collection The following operating data are collected via OPC interface: coal mining machine current, front scraper conveyor speed, and crusher temperature; real-time coordinates of the coal mining machine are collected via UWB positioning device; and monitoring images of the roadway conveyor are collected via industrial video camera.

[0072] All data is transmitted to edge computing nodes via Kafka message queues. During transmission, a CRC32 check mechanism is used, and data is automatically retransmitted if the check fails, ensuring the integrity of data collection. Special data (such as equipment maintenance records) is manually collected and uploaded through the software's manual import function, supporting Excel format files.

[0073] Step 3: Data Processing and Edge Preprocessing Data processing and edge preprocessing are key to reducing transmission pressure at the central end and improving data quality. Through standardized processing and local preprocessing, effective data can be screened and data formats optimized. Figure 3 Step 3 can be summarized as follows: real-time data clarity; feature extraction; edge compression.

[0074] 1) Standardization Processing (1) Data format conversion: Edge computing nodes convert data from different manufacturers' proprietary protocols into the JSON standard format. For example, the binary rotation speed data "0x000C" of the German DBT front scraper machine is converted into "{"Equipment code":"SG—03—140211006497","Measurement point code":"SG—03—140211006497—S01—001","Spin speed":1.2,"Unit":"m / s","Collection time":"2025—03—19 11:00:00"}", ensuring that the data format is consistent.

[0075] (2) Unit unification: The production data “300 tons / hour” was converted to “5000 kg / minute”, the temperature data “38℃” remained unchanged, and the current data “220A” remained unchanged to eliminate the influence of dimensional differences on subsequent analysis.

[0076] (3) Outlier filtering: Data exceeding the reasonable range is removed based on the 3σ principle. For example, belt conveyor speed > 3m / s or < 0.5m / s, crusher temperature > 80℃ or < -10℃, scraper conveyor speed 3.2m / s at a certain moment are judged as abnormal and filtered to avoid invalid data interfering with the analysis results.

[0077] 2) Edge preprocessing (1) Real-time filtering: The Kalman filter algorithm is used to process the coal mining machine current data, filter the current fluctuation noise, reduce the current data fluctuation amplitude from ±5A to ±1A, and improve the data accuracy.

[0078] (2) Local anomaly judgment: Set the equipment rated parameter threshold (e.g., the rated current of the coal mining machine is 216A). When the current is detected to be >216A, a preliminary warning is triggered locally. Only abnormal data is uploaded to the central end, while normal data (e.g., 200-216A) is stored locally, reducing the upload of invalid data and reducing the processing pressure on the central end.

[0079] Step 4: Linkage Analysis and Idle Speed ​​Identification Linkage analysis and idling identification are the core functions of this invention. Through GIS spatial mapping and idling model, it realizes the linkage verification of "data-space-video" and the identification of invalid equipment operation, thereby improving the efficiency of fault diagnosis and equipment performance. Figure 3 Step 4 can be summarized as follows: multi-dimensional data relationships; spatial feature matching; and anomaly pattern marking.

[0080] 1) GIS spatial coordinate mapping Edge computing nodes associate the real-time coordinates of the coal mining machine obtained by the UWB positioning device with the preset position information of the video camera, such as the preset position coordinates of the camera on the 30# support of the fully mechanized mining face, to establish the correspondence between the equipment position and the camera and realize automatic matching of video images.

[0081] When the equipment moves (for example, the coal mining machine advances 5 meters), the system automatically updates the equipment coordinates and associates them with the cameras in the new area, ensuring that the video footage always matches the equipment position, without the need for manual adjustment of the camera angle.

[0082] 2) "Data-Spatial-Video" Linked Verification For example, when the current of the front scraper conveyor is detected to be greater than 216A, the linkage analysis module completes three operations within 10 seconds: marks the location of the abnormal equipment in red on the GIS mine map; automatically calls the three video cameras of the 10# support, 15# support, and belt head in that area; captures the video clip at the moment of the abnormality and displays it synchronously with the current curve on the mine dispatch screen.

[0083] Dispatchers can intuitively view abnormal scenarios through the mine dispatching screen, quickly determine the cause of the fault, shorten the troubleshooting time, and improve processing efficiency.

[0084] 3) Idle running identification Construct an idling identification model: Invalid operation is determined by calculating the ratio of equipment running time to output (idling coefficient). Idle coefficient = (equipment running time - output / time consumed per unit output) / equipment running time. Time consumed per unit output is the time consumed per ton of coal when the equipment is running under normal load (for example, the time consumed per ton of coal when the coal mining machine is running under normal load is 0.002 hours / ton).

[0085] In the application scenario, if the coal mining machine runs for 9.95 hours on a given day, corresponding to an output of 4975 tons, the idling coefficient = (9.95 - 4975 × 0.002) / 9.95 = 0, which is considered as no idling. If the coal mining machine runs for 2 hours at a certain moment, with an output of only 500 tons (normally it should be 1000 tons), the idling coefficient = (2 - 500 × 0.002) / 2 = 0.5 > 0.3 (preset threshold), triggering an idling warning and pushing optimization suggestions such as "adjust the angle of the coal mining machine's cutting teeth to improve coal loading efficiency" to reduce idling energy consumption.

[0086] Step 5: Equipment Health Assessment Equipment health assessment is key to early warning of equipment failures and extending equipment lifespan. By comprehensively calculating the health index using multiple parameters, it provides a scientific basis for equipment maintenance. Figure 3 Step 5 can be summarized as follows: health indicator calculation; historical data comparison; risk level classification.

[0087] 1) Parameter Acquisition Collect operating parameters of the scraper conveyor in front of the fully mechanized coal mining machine, such as cumulative running time of 1800 hours, number of failures in the past 30 days (2 times, namely "chain jamming" and "motor overload"), and current fluctuation range of ±10%.

[0088] 2) Health Index Calculation The Analytic Hierarchy Process (AHP) is used to set parameter weights: for example, runtime accounts for 30%, number of failures accounts for 40%, and load fluctuation accounts for 30%.

[0089] 3) Calculate the scores for each parameter: For example, the maximum score for running time is 30 points, with 30 points for a cumulative running time of ≤2000 hours and 30 points for 1800 hours; the maximum score for the number of failures is 40 points, with 40 points for 0 failures and 5 points deducted for each additional failure, and 30 points for 2 failures; the maximum score for load fluctuation is 30 points, with 30 points for fluctuations ≤±10%, and 30 points for a load fluctuation of ±10% for this scraper conveyor.

[0090] If the health index is 30+30+30=90 points (out of 100), the health status is good and no immediate maintenance is required; if a crusher has accumulated 2500 hours of operation (running time score 25 points), has experienced 4 failures in the past 30 days (failure number score 20 points), and has a load fluctuation of ±18% (load fluctuation score 20 points), the health index is 25+20+20=65 points, which is close to the 60-point threshold, and a warning is sent: "It is recommended to check the crusher bearings in 15 days."

[0091] Step 6: Model Validation Model validation is crucial for verifying the accuracy of system functions. Through three simulations, it was ensured that the linkage analysis, idle detection, and health assessment functions meet the application requirements of fully mechanized mining faces. Figure 3 Step 6 can be summarized as follows: algorithm accuracy testing; real-world scenario verification; parameter optimization and adjustment.

[0092] 1) Verification of linkage analysis function Data from normal equipment in a fully mechanized coal mining face is collected. When the simulated current suddenly spikes to 220A (indicating an anomaly), the linkage analysis program is initiated.

[0093] If the GIS mine map annotation, video call, and data curve synchronous display are completed within 10 seconds, and the results are consistent with the actual abnormal scenario (e.g., the annotation location is between supports 10# and 15#, and the video shows the chain is stuck), it indicates that the linkage analysis function is accurate; if it takes too long or the results are deviated, the UWB positioning accuracy (e.g., whether it is affected by electromagnetic interference) or the video association logic needs to be checked, optimized, and then re-verified.

[0094] 2) Verification of idling recognition function Collect normal operating data of the coal mining machine (e.g., 2 hours of operation, output of 1000 tons, idling coefficient of 0) and idling simulation data (e.g., 2 hours of operation, output of 500 tons, idling coefficient of 0.5), and input them into the idling identification model respectively.

[0095] If normal data does not trigger an alert, but idle data does, and the alert result is consistent with the idle coefficient calculation, it indicates that the idle identification function is accurate. If there is a misjudgment or omission, the idle threshold (e.g., from 0.3 to 0.25) or the unit output time parameter needs to be adjusted to ensure identification accuracy.

[0096] 3) Health assessment function verification Select equipment with a known health status (e.g., a front scraper conveyor with a health index of 90 or a crusher with a health index of 65) and input the operating parameters into the health assessment module.

[0097] If the calculated health index is consistent with the known status (e.g., scraper conveyor 90 points, crusher 65 points) and the maintenance warning suggestions are reasonable (e.g., crusher suggests checking the bearings after 15 days), it indicates that the health assessment function is accurate; if the deviation is large, the AHP weight needs to be adjusted (e.g., increase the weight of the number of failures to 50%) and recalculate and verify.

[0098] In summary, the full-scale monitoring and spatial video linkage analysis system for six types of equipment in fully mechanized mining faces proposed in this application can effectively solve several key problems in the monitoring of these six types of equipment, providing strong technical support for efficient management and performance optimization of fully mechanized mining faces. These solutions include addressing data heterogeneity issues, achieving spatial-video linkage, accurately identifying idling status, and reducing data transmission pressure. Firstly, it addresses the data heterogeneity issue by using a unified coding standard to achieve unified identification of data from six types of equipment from different manufacturers. This improves data integration efficiency, shortens the time required for cross-equipment data analysis, and solves the problem of difficulty in summarizing and analyzing data from different manufacturers across equipment.

[0099] Secondly, it enables spatial-video linkage. By leveraging GIS mapping and automatic video matching functions, it can shorten the troubleshooting time, improve the accuracy of fault location, and enhance the efficiency and precision of troubleshooting.

[0100] Third, the accurate identification of idling status can distinguish between effective operation and idling status of equipment through the idling identification model, reduce the energy consumption of equipment idling, save electricity costs for each mine, and reduce equipment energy waste.

[0101] Fourth, reducing data transmission pressure can reduce invalid data uploads through edge preprocessing, shorten the processing delay at the central end, and at the same time, equipment health assessment can provide early warning of faults, reduce unplanned downtime, improve data processing efficiency, and ensure production continuity.

[0102] This application proposes a comprehensive monitoring and spatial-video linkage analysis system for the "six machines" (machines, machinery, and equipment) in a fully mechanized mining face. The system aims to provide a full-process, intelligent solution for monitoring the "six machines" in a fully mechanized mining face: by constructing a unified data coding standard, it breaks down data barriers between different manufacturers' equipment, enabling efficient aggregation and correlation analysis of cross-equipment data; relying on the deep linkage between GIS spatial positioning technology and video monitoring systems, it achieves automatic matching of abnormal equipment data with corresponding spatial locations and on-site videos, significantly shortening fault diagnosis time and improving emergency response efficiency; it introduces an intelligent idling identification algorithm to accurately distinguish between effective equipment operation and idling states, providing data support for energy consumption management and simultaneously pushing targeted efficiency optimization suggestions; it adds an edge-end data preprocessing module to filter invalid and duplicate data, reducing the processing pressure on the central end and lowering data transmission latency, ultimately helping the fully mechanized mining face achieve efficient management, energy consumption optimization, and safe and stable operation of the "six machines," meeting the high-quality equipment monitoring requirements of modern coal mine production.

[0103] The system proposed in this application is a system for full-data monitoring, spatial-video linkage verification, and idling identification of the "six machines" (machines, machinery, and equipment) in fully mechanized mining faces. It is particularly suitable for scenarios involving the full lifecycle monitoring, energy consumption optimization, fault early warning, and safety production control of the "six machines" in fully mechanized mining faces. By integrating standardized data processing, GIS spatial positioning, intelligent algorithm recognition, and edge computing technology, it forms a monitoring system covering the entire operating cycle of the "six machines." It not only addresses the technical pain points of each individual problem but also achieves deep integration of monitoring data, spatial location, and video footage through multi-module collaboration. This provides integrated technical support for the management of fully mechanized mining face equipment, from data acquisition and analysis to decision support, filling the current gap in the field of full-data monitoring of the "six machines."

[0104] Although preferred embodiments of the present application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of the embodiments of the present application.

[0105] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal device. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal device that includes said element.

[0106] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims. All of these forms are within the protection scope of this application.

Claims

1. A comprehensive monitoring and video-linked analysis system for six types of equipment in a fully mechanized mining face, characterized in that, include: Data acquisition module, data processing module, and linkage analysis module; The data acquisition module is used to collect full data from the six machines in the fully mechanized mining face. It uses a unified coding standard and OPC interface plus message queue technology to collect equipment operation data. The six machines include: coal mining machine, front scraper conveyor, rear scraper conveyor, transfer conveyor, crusher, and belt conveyor. The data processing module is used to standardize the collected operating data of the six machines and to align the real-time status data of the equipment with the video stream of the working face roadway using GIS spatial coordinate mapping technology. The linkage analysis module is used to dynamically mark the location of abnormal equipment on the mine map and automatically retrieve the corresponding video footage to form a linkage verification mechanism. At the same time, it constructs an idling identification model based on the equipment load curve and production data in the operation data, judges the invalid operation state of the equipment, and pushes efficiency optimization suggestions. The linkage verification mechanism refers to the mechanism of displaying an abnormal alarm based on its real-time operation data, spatial location, and related video when any of the six machines malfunctions.

2. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 1, characterized in that, The unified coding standard is used to achieve unified identification of equipment data from different manufacturers. It includes the unique identifier of each of the six machines and the coding rules for measurement points. The unique identifier is encoded in the format of device type code + region code + serial number, totaling 18 bits; The measurement point coding rule is: device code + measurement point type code + measurement point serial number.

3. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 1, characterized in that, The message queue of the data acquisition module uses a Kafka cluster. The Kafka cluster supports high-concurrency writing of various types of data and ensures data transmission integrity through a data verification mechanism, including CRC32 checksum.

4. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 1, characterized in that, The standardization processes of the data processing module include: data format conversion, unit standardization, and outlier filtering; The data format is converted to convert proprietary protocol data from different manufacturers into JSON format; The units mentioned are uniformly converted from tons per hour to kilograms per minute; The outlier filtering is based on the 3σ principle to remove data that exceeds a reasonable range.

5. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 1, characterized in that, The GIS spatial coordinate mapping technology is as follows: real-time coordinates are obtained through the UWB positioning module built into the device, with a positioning accuracy of ≤0.5 meters; The GIS spatial coordinate mapping technology is associated with the preset position information of the video surveillance cameras to automatically match the positions of each of the six devices with the video footage.

6. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 1, characterized in that, When the linkage verification mechanism detects an abnormal device current, it completes the following operations within 10 seconds: The location of abnormal equipment is highlighted on the GIS mining map, and more than three video cameras in the area are automatically called. Video clips of the abnormal moment are captured and displayed synchronously with the data curve, which is the curve corresponding to the operating data of the six abnormal machines.

7. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 1, characterized in that, The idling identification model determines the ineffective operating status of equipment by calculating the idling coefficient of each of the six machines. The idling coefficient is the ratio of operating time to output and is related to the load correction coefficient. When the idling coefficient exceeds a preset threshold, the six machines exceeding the preset threshold are determined to be in an invalid operating state.

8. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 1, characterized in that, The performance optimization suggestions include: equipment parameter adjustment, start-up and shutdown sequence optimization, and maintenance cycle reminders; The equipment parameter adjustments include, but are not limited to: reducing the crusher frequency from 50Hz to 45Hz; The optimization of start-stop timing includes, but is not limited to: the scraper conveyor starts 30 seconds later than the coal mining machine; The maintenance cycle reminder includes, but is not limited to: it is recommended to change the lubricating oil of the transfer machine every 15 days.

9. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 1, characterized in that, The data processing module supports the deployment of edge computing nodes; The edge computing nodes are deployed in control cabinets near the work surface to complete data preprocessing. The data preprocessing reduces invalid data uploaded to the central end. The data preprocessing includes: real-time filtering and local anomaly detection.

10. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 1, characterized in that, Also includes: Equipment health assessment module; The equipment health assessment module calculates the health index of each six-machine device based on the cumulative running time, number of failures, and load fluctuation range parameters of each six-machine device using the analytic hierarchy process. The equipment health index ranges from 0 to 100 points. When the equipment health index of any six-machine device is lower than 60 points, a maintenance warning is automatically triggered.

11. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 1, characterized in that, The dynamic annotation of the mining map in the linkage analysis module supports multi-layer display, and the dynamic annotation of the mining map can trace back the status and video footage of each of the six machines at any time through the time axis. The multi-layer includes: equipment location layer, data anomaly layer, and video overlay layer.

12. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 7, characterized in that, The preset threshold used in the idling identification model is a dynamically adjustable judgment threshold. The method for obtaining the determination threshold by the idle recognition model includes: The coal seam hardness, moisture content, and dip angle are collected using a data acquisition unit deployed on the working face support. The data acquisition unit includes a coal seam hardness sensor and a moisture content sensor. A weighted Naive Bayes fusion algorithm is used to convert the coal seam hardness, moisture content, and dip angle into load correction coefficients. Each load correction coefficient is equal to the sum of the coal seam hardness correction value, moisture content correction value, and dip angle correction value, divided by 3. Specifically: when the coal seam hardness is less than or equal to 2 MPa, the hardness correction value is 1.0; when it is between 2 and 4 MPa, the hardness correction value is 0.8; and when it is greater than 4 MPa, the hardness correction value is 0.

6. Similarly, when the moisture content is less than or equal to 8%, the moisture content correction value is 1.0; when it is between 8% and 15%, the moisture content correction value is 0.9; and when it is greater than 15%, the moisture content correction value is 0.

7. Finally, when the dip angle is less than or equal to 15°, the dip angle correction value is 1.0; when it is between 15° and 30°, the dip angle correction value is 0.85; and when it is greater than 30°, the dip angle correction value is 0.

75. The judgment threshold is dynamically adjusted according to the equipment type and load correction coefficient. Specifically, for a coal mining machine, the judgment threshold is 0.3 when the load correction coefficient is equal to 1.0, and is increased to 0.45 when the correction coefficient is equal to 0.6; for a scraper conveyor, the judgment threshold is 0.28 when the load correction coefficient is equal to 1.0, and is increased to 0.38 when the correction coefficient is equal to 0.

75. The idling coefficient is equal to the ratio of the equipment's running time to its output. When the idling coefficient is multiplied by the load correction coefficient, and the idling coefficient exceeds the corresponding judgment threshold, the six machines that exceed the judgment threshold are determined to be in an invalid operating state.

13. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 6, characterized in that, When the linkage verification mechanism detects an abnormal equipment current, it uses a spatiotemporal tracing module to highlight the location of the abnormal equipment on the GIS mining map, automatically call more than 3 video cameras in the area, capture video clips at the time of the abnormality, and display them synchronously with the data curve. The spatiotemporal tracing module includes: an equipment correlation modeling unit, a spatiotemporal causal calculation unit, and a tracing result visualization unit. The equipment association modeling unit constructs a spatiotemporal association matrix based on the physical connection relationship and historical fault data of the six equipments, and marks the influence delay and influence intensity between each equipment. The spatiotemporal causal calculation unit uses a sliding time window + causal entropy algorithm to extract the running data from 60 seconds before the abnormality to 30 seconds after the abnormality when any device malfunctions, calculate the causal entropy value of each device abnormality event, and determine the device with a causal entropy value less than 0.3 as the root cause device. The source tracing result visualization unit marks the root cause device with red flashing, the primary influencing device with orange solid light, and the secondary influencing device with yellow solid light on the GIS mining map. It also generates an anomaly propagation timeline to show the time of anomaly occurrence, impact delay, and data change trend of each device, and synchronously displays video clips and data curves at the time of anomalies.

14. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 9, characterized in that, The edge computing node includes: an edge-end intelligent preprocessing system based on time-series prediction; The edge-end intelligent preprocessing system includes: a time-series data prediction unit, a predictive filtering unit, and a local early warning push unit; The time-series data prediction unit uses an LSTM + attention mechanism to build a lightweight prediction model. Based on the operating data of each device over the past 5 minutes, it predicts the data flow trend for the next 1 minute and outputs the predicted value and confidence interval. The lightweight prediction model uses int8 precision quantization compression to adapt to edge industrial control computers, and the data processing time for a single device is less than or equal to 50ms. The predictive filtering unit follows the following rules: if the predicted value deviates from the current data by less than 5% and the predicted trend is returning to normal, the locally stored data is not uploaded to the central end; if the predicted value exceeds the normal range, the predicted abnormal data and the current data are packaged and uploaded to the central end. The local early warning push unit triggers a local audible and visual early warning when an anomaly is predicted.

15. The mine fully mechanized mining face six-machine equipment full-scale monitoring and video linkage analysis system according to claim 10, characterized in that, The equipment health assessment module is a dynamic health assessment module with two dimensions: equipment type and operating condition. It includes: equipment type weight library, operating condition level classification unit, dynamic weight calculation unit and health index correction unit. The equipment type weight library sets differentiated basic weights for the six types of equipment, including: coal mining machine: 25% for running time, 35% for failure frequency, 20% for load fluctuation, and 20% for cutter tooth wear; crusher: 20% for running time, 40% for failure frequency, 15% for load fluctuation, and 25% for bearing temperature; belt conveyor: 30% for running time, 35% for failure frequency, 25% for load fluctuation, and 10% for belt tension. The operating condition classification unit divides each device into three levels based on its real-time load rate: low load (load rate < 60%), medium load (60% ≤ load rate ≤ 90%), and high load (load rate > 90%). The adjustment rules for the dynamic weight calculation unit are as follows: under high load conditions, the weight of load fluctuation is increased by 20% and the weight of failure frequency is increased by 10%; under low load conditions, the weight of running time is increased by 15% and the weight of newly added core parameters is increased by 10%. The core parameters include: the wear degree of the cutting teeth, the bearing temperature, and the belt tension. The health index correction unit introduces a working condition influence coefficient, setting 1.0 for low load, 1.05 for medium load, and 1.1 for high load. The final health index = basic health index × working condition influence coefficient, and the basic health index is calculated based on the adjusted weights.

Images