Mine slope safety monitoring system and method based on unmanned aerial vehicle

Abstract

The invention relates to the technical field of mine slope monitoring, in particular to a mine slope safety monitoring system and method based on an unmanned aerial vehicle, and the method comprises the following steps: S1, formulating a detailed preset route according to a slope range and risk points in a mining area, setting key observation points and alternative routes, system detection is carried out on the carried various sensors and the unmanned aerial vehicle platform; and S2, the unmanned aerial vehicle platform autonomously takes off according to a preset route and executes a flight task, and all-directional and multi-angle data acquisition is performed on the mine slope by using a carried high-resolution optical camera, a laser radar, an infrared thermal imager and a GNSS / IMU combined positioning system. And automatic preprocessing, feature extraction and risk assessment are realized at the ground center, so that the intelligence, real-time performance and precision of mine slope monitoring are remarkably improved, the problems of low efficiency and slow response of a traditional means are solved, and risk grading and an automatic response mechanism are carried out based on an intelligent algorithm.

Description

Technical Field

[0001] The present invention relates to the technical field of mine slope monitoring, and particularly to a mine slope safety monitoring system and method based on an unmanned aerial vehicle (UAV). Background Art

[0002] A mine slope safety monitoring system based on an unmanned aerial vehicle is an intelligent system that uses UAV technology to monitor mine slopes in real time and efficiently;

[0003] This system uses a UAV equipped with high-precision sensors (such as lidar, camera, infrared thermal imager, etc.) to regularly inspect the mine slope, collect key data such as the terrain, cracks, and displacements of the slope, and the collected data is transmitted back to the ground control center in real time through wireless transmission technology. Combining with a geographic information system (GIS) and data analysis software, the stability of the slope is evaluated and warned. This system can timely detect potential safety hazards, provide a scientific basis for mine safety management, effectively reduce the risk of accidents such as mine slope collapses, and ensure the safe and stable operation of mine production;

[0004] In the existing process of mine slope safety monitoring based on UAVs, the dimension of data acquisition is single, which will lead to low accuracy of mine slope monitoring. And although some systems have a risk prompt function, they lack a risk level division and automatic response mechanism based on intelligent algorithms, and cannot achieve accurate early warning and rapid linkage. Therefore, a mine slope safety monitoring system and method based on UAVs are proposed for the above problems. Summary of the Invention

[0005] The purpose of the present invention is to provide a mine slope safety monitoring system and method based on UAVs to solve the problems in the existing process of mine slope safety monitoring based on UAVs, where the dimension of data acquisition is single, which will lead to low accuracy of mine slope monitoring, and although some systems have a risk prompt function, they lack a risk level division and automatic response mechanism based on intelligent algorithms, and cannot achieve accurate early warning and rapid linkage.

[0006] To achieve the above purpose, the present invention provides the following technical solutions:

[0007] A mine slope safety monitoring system and method based on UAVs includes the following steps:

[0008] S1: In the mining area, a detailed preset flight path is formulated according to the slope range and risk points, key observation points and alternative flight paths are set, and a system detection is carried out on the various sensors and UAV platforms carried;

[0009] S2: The UAV platform takes off autonomously according to the preset flight path and executes the flight mission. It uses the high-resolution optical camera, lidar, infrared thermal imager, and GNSS / IMU integrated positioning system carried on it to collect all-round and multi-angle data of the mine slope, while realizing autonomous obstacle avoidance and dynamic flight path adjustment.

[0010] S3: Use a high-speed wireless communication link to transmit various data collected by the UAV to the ground control center in real time. The transmitted data adopts a data packet structure, where each data packet P contains: image data I, point cloud data L, temperature data T, positioning data G, sensor status data S, and an additional cyclic redundancy check code E.

[0011] S4: Preprocess, extract features, and conduct safety assessment on the data received at the ground control center.

[0012] S5: For the areas where potential safety hazards are detected, a risk grading algorithm based on fuzzy logic is used for early warning, and its risk degree R d The calculation formula is:

[0013] R d =μ low (S)·w low +μ medium (S)·w medium +μ high (S)·w high

[0014] In the formula, μ low (S) is the low-risk membership function, defined as: μ medium (S) is the medium-risk membership function, defined as: μ high (S) is the high-risk membership function, defined as: S is the comprehensive safety assessment value, S th is the safety assessment threshold, S min is the lowest reference value of the safety assessment function, k low 、k medium and k high are the fuzzification constants, w low 、w medium and w high are the weights corresponding to low, medium, and high risks;

[0015] Based on the calculated risk degree R d , the system automatically marks the risk area on the display platform, and sends a warning signal to relevant personnel via text message, email, or a dedicated alarm system, while triggering the emergency response process.

[0016] As a further optimization of the present invention, it further includes the following steps:

[0017] S6: Store all monitoring data, processing results and warning records in the database and provide a data query interface.

[0018] As a further optimization of the present invention, the cyclic redundancy check code E is used to verify the integrity of data transmission, and the calculation formula of the cyclic redundancy check code E is:

[0019] E = CRC(P)

[0020] In the formula, CRC(·) is the cyclic redundancy check function.

[0021] As a further optimization of the present invention, the data preprocessing in step four includes:

[0022] Preprocessing of image data I: Use Gaussian filtering to eliminate noise interference, and the calculation formula is:

[0023]

[0024] Obtain the preprocessed image data X for subsequent use by the convolutional neural network model. In the formula, σ is the standard deviation of Gaussian filtering;

[0025] Point cloud data P orig Preprocessing: Use the spatial calibration matrix M to perform spatial alignment on the original three-dimensional point data, and the calculation formula is:

[0026] P' = M·P orig

[0027] Obtain the corrected point cloud data P'. In the formula, P orig is the original point cloud data.

[0028] As a further optimization of the present invention, the feature extraction in step four includes:

[0029] Image feature extraction: Use a convolutional neural network to detect cracks, color changes, and abnormal textures in the image, and the calculation formula is:

[0030] Y = f CNN (X,θ)

[0031] In the formula, Y is the feature extraction result, θ is the parameter of the convolutional neural network model, and f CNN (·) is the convolutional neural network feature extraction function;

[0032] Point cloud reconstruction and deformation recognition:

[0033] Use the multi-view geometry algorithm to reconstruct the corrected point cloud P' into a three-dimensional model R, and the calculation formula is:

[0034] R = f recon (P′, M)

[0035] Wherein, f recon (·) is a multi - perspective 3D reconstruction function.

[0036] As a further optimized content of the present invention, wherein: the safety assessment in step four includes:

[0037] Define a comprehensive safety assessment function S:

[0038] S = α·f crack (Y)+β·f slope (R)+γ·f deformation (R)

[0039] Wherein, f crack (Y) is a crack risk function based on image features, f slope (R) is a slope analysis function based on a 3D model, f deformation (R) is a 3D deformation amount extraction function, and α, β, and γ are artificially set risk weight coefficients;

[0040] Judgment logic:

[0041] If S < S th , then there is a safety hazard in the current slope, enter the early warning system, and perform fuzzy logic risk level judgment and alarm mechanism.

[0042] As a further optimized content of the present invention, wherein: it includes a drone platform, a sensor component, a ground control center, a wireless communication module, and a data processing platform, wherein:

[0043] The drone platform has functions of autonomous flight, path planning, and obstacle avoidance, and is equipped with a high - resolution optical camera, a lidar, an infrared thermal imager, and a GNSS / IMU integrated positioning system;

[0044] The sensor component is used to collect image data, point cloud data, temperature data, and positioning data;

[0045] The wireless communication module is used to transmit various types of monitored data collected in real - time to the ground control center;

[0046] The ground control center integrates a data pre - processing module, a feature extraction module, a safety assessment module, and a risk early - warning module;

[0047] The system further includes a database module, which is used to store historical monitoring data and risk early - warning records for a long time and provide a query interface.

[0048] As a further optimized content of the present invention, wherein: the data processing platform of the ground control center further includes:

[0049] An image data preprocessing unit: used to perform image noise elimination, enhancement, and normalization processing;

[0050] A point cloud calibration unit: to perform spatial coordinate calibration on the original three-dimensional point cloud data;

[0051] A feature recognition unit: to extract crack and texture anomaly features in the image through a deep learning model, and perform three-dimensional reconstruction and deformation analysis;

[0052] A safety assessment unit: to perform comprehensive analysis by combining multi-source data and output the slope stability assessment result;

[0053] A risk grading and warning unit: to automatically divide the risk level based on the assessment result, trigger the sound and light, SMS or platform warning mechanism, and at the same time push it to the emergency response module for response processing.

[0054] Compared with the prior art, the beneficial effects of the present invention are:

[0055] 1. In the present invention, multi-source data is collected through the autonomous flight of the unmanned aerial vehicle, and automatic preprocessing, feature extraction, and risk assessment are realized at the ground center, significantly improving the intelligence, real-time performance, and accuracy of mine slope monitoring, solving the problems of low efficiency, poor accuracy, and slow response of traditional means, and based on the intelligent algorithm for risk level division and automatic response mechanism, accurate early warning and rapid linkage can be achieved;

[0056] 2. In the present invention, by constructing an integrated data management and processing system at the ground control center, not only the unified storage and query of monitoring data, processing results, and warning information are realized, but also the accuracy and real-time performance of slope hidden danger identification are significantly improved through an efficient data preprocessing and feature extraction process. Especially the introduction of technologies such as convolutional neural network, point cloud three-dimensional reconstruction, and multi-view analysis enables the automatic identification of risk features such as surface cracks, deformations, and temperature differences on the slope, reducing the manual identification error and improving the intelligent level of the overall monitoring system;

[0057] 3. In the present invention, the provided system architecture has high integration and automatic response capabilities. Through the collaborative work of the autonomous flight of the unmanned aerial vehicle and multi-source sensors, the dynamic perception of the slope environment is realized; at the same time, the system has a flexible risk warning mechanism, which can perform fuzzy logic risk level judgment based on multi-source data and automatically trigger multi-channel warnings such as sound and light and SMS, effectively improving the emergency response efficiency. In addition, a cyclic redundancy check mechanism is adopted to ensure the integrity and reliability of data transmission, providing security guarantees for data communication in high-risk environments. Description of the Drawings

[0058] Figure 1 This is a flowchart of a method for monitoring the safety of mine slopes based on an unmanned aerial vehicle (UAV) according to the present invention;

[0059] Figure 2 This is a system block diagram of a system for monitoring the safety of mine slopes based on an unmanned aerial vehicle (UAV) according to the present invention. Detailed implementation manners

[0060] Please refer to Figure 1-2 , the present invention provides a technical solution:

[0061] A system and method for monitoring the safety of mine slopes based on an unmanned aerial vehicle (UAV), comprising the following steps:

[0062] S1: In the mining area, a detailed preset flight path is formulated according to the slope range and risk points, key observation points and alternative flight paths are set, and various sensors and the UAV platform carried are subjected to system detection;

[0063] S2: The UAV platform takes off autonomously according to the preset flight path and executes the flight mission, uses the high-resolution optical camera, lidar, infrared thermal imager and GNSS / IMU combined positioning system carried to collect all-round and multi-angle data of the mine slope, and at the same time realizes autonomous obstacle avoidance and dynamic flight path adjustment;

[0064] S3: Use a high-speed wireless communication link to transmit various data collected by the UAV to the ground control center in real time. The transmitted data adopts a data packet structure, and each data packet P includes: image data I, point cloud data L, temperature data T, positioning data G, sensor status data S and an additional cyclic redundancy check code E;

[0065] S4: Preprocess, extract features and perform safety assessment on the received data at the ground control center;

[0066] S5: For the areas where potential safety hazards are detected, a risk grading algorithm based on fuzzy logic is used for early warning, and its risk degree R d The calculation formula is:

[0067] R d = μ low (S)·w low + μ medium (S)·w medium + μ high (S)·w high

[0068] In the formula, μ low (S) is the low-risk membership function, defined as: μ medium (S) is the medium-risk membership function, defined as: μ high(S) is the high-risk membership function, defined as: S is the comprehensive safety assessment value, S th is the safety assessment threshold, S min is the lowest reference value of the safety assessment function, k low 、k medium and k high are the fuzzification constants, w low 、w medium and w high are the weights corresponding to low, medium, and high risks;

[0069] Based on the calculated risk degree R d , the system automatically marks the risk area on the display platform, and sends early warning signals to relevant personnel via text messages, emails, or a dedicated alarm system. At the same time, it triggers the emergency response process, combines multi-type sensors and unmanned aerial vehicle autonomous flight technology, realizes high-precision, full-coverage, and multi-angle monitoring of the mine slope, effectively improves the monitoring efficiency and inspection safety, and is especially suitable for dangerous areas or areas that are difficult for humans to reach.

[0070] As a further implementation technical solution of this scheme, it also includes the following steps:

[0071] S6: Store all monitoring data, processing results, and early warning records in the database, and provide a data query interface, realizing the continuous retention and traceability function of the data, facilitating the comparative analysis and trend judgment of historical data, and supporting the risk tracing and responsibility definition of regulatory agencies and mining enterprises;

[0072] As a further implementation technical solution of this scheme, the cyclic redundancy check code E is used to verify the integrity of data transmission, and the calculation formula of the cyclic redundancy check code E is:

[0073] E = CRC(P)

[0074] In the formula, CRC(·) is the cyclic redundancy check function, which enhances the anti-interference ability and reliability of the system during wireless data transmission, ensures that key monitoring data is not lost or tampered with, and provides a stable data basis for subsequent processing;

[0075] As a further implementation technical solution of this scheme, the data preprocessing in step four includes:

[0076] Image data I preprocessing: Use Gaussian filtering to eliminate noise interference, and the calculation formula is:

[0077]

[0078] Obtain the preprocessed image data X for subsequent use in the convolutional neural network model. In the formula, σ is the standard deviation of Gaussian filtering;

[0079] Point cloud data P orig Preprocessing: Use the spatial calibration matrix M to perform spatial alignment on the original three-dimensional point data. The calculation formula is:

[0080] P′ = M·P orig

[0081] Obtain the corrected point cloud data P′, where P orig is the original point cloud data. Through the preprocessing of images and point cloud data, the recognition accuracy of subsequent models can be effectively improved, invalid noise can be filtered, and the overall recognition ability and stability of the monitoring system can be enhanced;

[0082] As a further technical solution for the implementation of this scheme, the feature extraction in step four includes:

[0083] Image feature extraction: Use a convolutional neural network to detect cracks, color changes, and abnormal textures in the image. The calculation formula is:

[0084] Y = f CNN (X,θ)

[0085] where Y is the feature extraction result, θ is the convolutional neural network model parameter, and f CNN (·) is the convolutional neural network feature extraction function;

[0086] Point cloud reconstruction and deformation recognition:

[0087] Use the multi-view geometry algorithm to reconstruct the corrected point cloud P′ into a three-dimensional model R. The calculation formula is:

[0088] R = f recon (P′,M)

[0089] where f recon (·) is the multi-view three-dimensional reconstruction function, which integrates artificial intelligence feature extraction and three-dimensional modeling technologies, can accurately identify abnormal slope structures, helps to detect potential collapse risks at an early stage, and improves the timeliness and accuracy of early warnings;

[0090] As a further technical solution for the implementation of this scheme, the safety assessment in step four includes:

[0091] Define the comprehensive safety assessment function S:

[0092] S = α·f crack (Y)+β·f slope (R)+γ·f deformation (R)

[0093] where f crack (Y) is the crack risk function based on image features, f slope(R) is a slope analysis function based on a 3D model, f deformation (R) is a 3D deformation extraction function, and α, β, and γ are artificially set risk weight coefficients;

[0094] Judgment logic:

[0095] If S < S th , then there are potential safety hazards in the current slope, enter the early warning system, perform fuzzy logic risk level judgment and alarm mechanism, and conduct quantitative safety assessment by integrating multi-dimensional data sources, effectively avoiding misjudgments caused by traditional single-index assessments. The fuzzy logic grading mechanism enhances the system's adaptability to complex situations;

[0096] As a further implementation technical solution of this scheme, it includes a drone platform, a sensor component, a ground control center, a wireless communication module, and a data processing platform, where:

[0097] The drone platform has functions of autonomous flight, path planning, and obstacle avoidance, and is equipped with a high-resolution optical camera, a lidar, an infrared thermal imager, and a GNSS / IMU combined positioning system;

[0098] The sensor component is used to collect image data, point cloud data, temperature data, and positioning data;

[0099] The wireless communication module is used to transmit various types of monitored data collected in real time to the ground control center;

[0100] The ground control center integrates a data preprocessing module, a feature extraction module, a safety assessment module, and a risk warning module;

[0101] The system also includes a database module, which is used to store historical monitoring data and risk warning records for a long time and provide query interfaces. The system has a highly modular, integrated, and intelligent hardware platform, is suitable for various complex mine environments, supports flexible deployment and expansion, and overall improves the intelligent level of slope monitoring operations;

[0102] As a further implementation technical solution of this scheme, the data processing platform of the ground control center further includes:

[0103] Image data preprocessing unit: used to perform image noise elimination, enhancement, and standardization processing;

[0104] Point cloud calibration unit: perform spatial coordinate calibration on the original 3D point cloud data;

[0105] Feature recognition unit: extract crack and texture anomaly features in the image through a deep learning model, and perform 3D reconstruction and deformation analysis;

[0106] Safety assessment unit: Conduct comprehensive analysis by integrating multi-source data and output the evaluation results of slope stability;

[0107] Risk classification and early warning unit: Automatically classify the risk levels based on the evaluation results, trigger the sound and light, SMS or platform early warning mechanism, and at the same time push them to the emergency response module for response processing. Through the refined processing module, the ground control center is equipped with the ability of rapid response and hierarchical processing, improving the monitoring efficiency and decision-making intelligence, and strengthening the stability and expandability of the system in actual application.

[0108] In this article, specific examples are used to elaborate on the principle and implementation mode of the present invention. The description of the above examples is only used to help understand the method and its core idea of the present invention. The above is only the preferred implementation mode of the present invention. It should be noted that due to the limitation of literal expression, and objectively there are infinite specific structures. For those of ordinary skill in the art in this technical field, without departing from the principle of the present invention, several improvements, refinements or changes can also be made, or the above technical features can be combined in an appropriate way; these improvements, refinements, changes or combinations, or directly applying the concept and technical solution of the invention to other occasions without improvement, should all be regarded as the protection scope of the present invention.

Claims

1. A method for safety monitoring of mine slopes based on drones, characterized in that, It includes the following steps: S1: In the mining area, formulate a detailed preset flight path according to the slope range and risk points, set key observation points and alternative flight paths, and conduct system detection on various sensors and the UAV platform carried; S2: The UAV platform takes off autonomously according to the preset flight path and executes the flight mission. Use the high-resolution optical camera, lidar, infrared thermal imager and GNSS / IMU combined positioning system carried to collect all-round and multi-angle data of the mine slope, and at the same time realize autonomous obstacle avoidance and dynamic flight path adjustment; S3: Use a high-speed wireless communication link to transmit various data collected by the UAV to the ground control center in real time. The transmitted data adopts a data packet structure, where each data packet P contains: image data I, point cloud data L, temperature data T, positioning data G, sensor status data S and an additional cyclic redundancy check code E; S4: Preprocess, extract features and conduct safety assessment on the received data at the ground control center; S5: For the areas where potential safety hazards are detected, a risk grading algorithm based on fuzzy logic is used for early warning, and its risk degree R d is calculated by the following formula: R d = μ low (S)·w low + μ medium (S)·w medium + μ high (S)·w high where μ low (S) is the low - risk membership function, defined as: μ medium (S) is the medium - risk membership function, defined as: μ high (S) is the high - risk membership function, defined as: S is the comprehensive safety assessment value, S th is the safety assessment threshold, S min is the lowest reference value of the safety assessment function, k low 、k medium and k high are fuzzification constants, w low 、w medium and w high are the weights corresponding to low, medium, and high risks; Based on the calculated risk level R d , the system automatically marks the risk area on the display platform, sends early warning signals to relevant personnel via text message, email or a dedicated alarm system, and simultaneously triggers the emergency response process.

2. The method for monitoring the safety of a mine slope based on an unmanned aerial vehicle according to claim 1, wherein: It also includes the following steps: S6: Store all monitoring data, processing results and early warning records in the database and provide a data query interface.

3. A method for safety monitoring of mine slopes based on drones according to claim 1, characterized in that: The cyclic redundancy check code E is used to verify the integrity of data transmission. The calculation formula of the cyclic redundancy check code E is: E = CRC(P) In the formula, CRC(·) is the cyclic redundancy check function.

4. The method for monitoring the safety of a mine slope based on an unmanned aerial vehicle according to claim 1, wherein: The data preprocessing in step four includes: Preprocessing of image data I: Use Gaussian filtering to eliminate noise interference. The calculation formula is: Obtain the preprocessed image data X for use in the subsequent convolutional neural network model. In the formula, σ is the standard deviation of Gaussian filtering; Point cloud data P orig Preprocessing: Use the spatial calibration matrix M to perform spatial alignment on the original three-dimensional point data. The calculation formula is as follows: P′ = M·P orig Obtain the corrected point cloud data P′, where P orig is the original point cloud data.

5. A method for safety monitoring of mine slopes based on drones according to claim 1, characterized in that: The feature extraction in step four includes: Image feature extraction: Use a convolutional neural network to detect cracks, color changes, and abnormal textures in the image. The calculation formula is: Y = f CNN (X, θ) where Y is the feature extraction result, θ is the parameter of the convolutional neural network model, and f CNN (·) is the feature extraction function of the convolutional neural network; Point cloud reconstruction and deformation identification: Use a multi-view geometry algorithm to reconstruct the corrected point cloud P′ into a three-dimensional model R. The calculation formula is: R = f recon (P′, M) where f recon (·) is a multi-view 3D reconstruction function.

6. The method for monitoring the safety of a mine slope based on a drone according to claim 1, wherein: The safety assessment in step four includes: Define a comprehensive safety assessment function S: S = α·f crack (Y) + β·f slope (R) + γ·f deformation (R) where, f crack (Y) is the crack risk function based on image features, f slope (R) is the slope analysis function based on the 3D model, f deformation (R) is the 3D deformation amount extraction function, and α, β, and γ are risk weight coefficients set artificially; Judgment logic: If S < S th , there are potential safety hazards in the current slope, and it enters the warning system for fuzzy logic risk level judgment and alarm mechanism.

7. The mine slope safety monitoring system based on unmanned aerial vehicle according to claim 1, wherein: It includes a UAV platform, a sensor component, a ground control center, a wireless communication module and a data processing platform, where: The UAV platform has functions of autonomous flight, path planning and obstacle avoidance, and is equipped with a high-resolution optical camera, lidar, infrared thermal imager and GNSS / IMU combined positioning system; The sensor component is used to collect image data, point cloud data, temperature data and positioning data; The wireless communication module is used to transmit various monitoring data collected to the ground control center in real time; The ground control center integrates a data preprocessing module, a feature extraction module, a safety assessment module and a risk early warning module; The system also includes a database module for long-term storage of historical monitoring data and risk early warning records and providing a queryable interface.

8. The mine slope safety monitoring system based on an unmanned aerial vehicle according to claim 1, wherein: The data processing platform of the ground control center further includes: Image data preprocessing unit: Used to perform image noise elimination, enhancement and standardization processing; Point cloud calibration unit: Calibrate the spatial coordinates of the original three-dimensional point cloud data; Feature recognition unit: Extract crack and texture abnormal features in the image through a deep learning model, and conduct three-dimensional reconstruction and deformation analysis; Safety assessment unit: Conduct comprehensive analysis by combining multi-source data and output the assessment results of slope stability; Risk classification and early warning unit: Automatically classify risk levels based on the assessment results, trigger the audible and visual, SMS or platform early warning mechanism, and at the same time push it to the emergency response module for response processing.

Images