Roadway surrounding rock comprehensive monitoring and early warning method and device

Abstract

The invention discloses a comprehensive monitoring and early warning method and device for roadway surrounding rock, and relates to the technical field of underground engineering geotechnical engineering. Collecting physical parameter data of the surrounding rock in real time through a sensor network; transmitting the collected physical parameter data to a monitoring center through a network, and carrying out filtering, denoising and abnormal value elimination to obtain standard multi-source data; comprehensively analyzing the multi-source data by using a method of combining Bayesian estimation and fuzzy logic, and extracting a stability score of the surrounding rock stability; and training the relationship between the physical parameter data and the stability score, establishing a prediction model of the surrounding rock stability, and performing early warning when the predicted real-time stability score exceeds an early warning threshold. Therefore, through the constructed surrounding rock stability prediction model, comprehensive and real-time monitoring and high-precision early warning of the surrounding rock of the roadway are realized, the unstable condition of the surrounding rock is found and processed in time, and safety accidents are avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of underground geotechnical engineering, and in particular to a method and device for comprehensive monitoring and early warning of surrounding rocks of a tunnel. Background Art

[0002] In underground mining and tunnel construction, the stability of the surrounding rock mass plays a critical role in project safety and progress. Currently, traditional monitoring methods rely primarily on manual inspections and the measurement of single parameters, such as displacement and stress. These methods suffer from limited monitoring parameters, incomplete information, insufficient real-time performance, and imperfect early warning levels. These methods struggle to accurately and timely reflect the true state of the surrounding rock mass, and therefore fail to meet the stringent safety management requirements of modern projects.

[0003] With the development of sensor, communication, and data processing technologies, multi-parameter integrated monitoring and intelligent early warning methods have gradually become a research hotspot. However, existing technologies still have shortcomings in sensor placement, data transmission reliability, data preprocessing and fusion, and prediction model accuracy. There is an urgent need for a method that can comprehensively and real-timely monitor the surrounding rock conditions of tunnels and provide high-precision early warning capabilities to improve the safety management level of underground projects. Summary of the Invention

[0004] The present invention aims to solve one of the technical problems in the related art at least to a certain extent.

[0005] To this end, the first purpose of the present invention is to propose a comprehensive monitoring and early warning method for tunnel surrounding rocks. By constructing a prediction model of surrounding rock stability, comprehensive, real-time monitoring and high-precision early warning of tunnel surrounding rocks are achieved, and unstable conditions of surrounding rocks are discovered and handled in a timely manner, avoiding the occurrence of safety accidents.

[0006] The second purpose of the present invention is to provide a comprehensive monitoring and early warning device for tunnel surrounding rocks.

[0007] A third object of the present invention is to provide an electronic device.

[0008] A fourth object of the present invention is to provide a non-transitory computer-readable storage medium storing computer instructions.

[0009] To achieve the above objectives, a first embodiment of the present invention provides a method for comprehensive monitoring and early warning of surrounding rock in a roadway, comprising the following steps:

[0010] Construct a sensor network for the tunnel surrounding rock. The sensor network consists of multiple types of sensors deployed at multiple characteristic monitoring locations in the tunnel. The characteristic monitoring locations are determined by the geological structure of the tunnel and the geological characteristics of the surrounding rock. The multiple types of sensors include displacement sensors, stress sensors, acceleration sensors, temperature sensors, and humidity sensors.

[0011] The sensor network collects physical parameter data of the surrounding rock at each characteristic monitoring position in real time, wherein the physical parameter data includes displacement change data of the surrounding rock collected by the displacement sensor, stress state data of the surrounding rock collected by the stress sensor, vibration characteristic data of the surrounding rock collected by the acceleration sensor, and environmental parameters in the tunnel collected by the temperature sensor and the humidity sensor;

[0012] Transmitting the displacement change data, stress state data, vibration characteristic data, and environmental parameters to a preset monitoring center via a network;

[0013] The monitoring center filters, removes noise and eliminates outliers on displacement change data, stress state data, vibration characteristic data and environmental parameters to obtain standard multi-source data;

[0014] By combining Bayesian estimation and fuzzy logic, the multi-source data are comprehensively analyzed to extract the stability score of the surrounding rock stability;

[0015] Taking physical parameter data as input and the stability score of surrounding rock stability as output, a prediction model of surrounding rock stability is established through training based on machine learning algorithm;

[0016] Set a warning threshold. When the real-time stability score predicted by the prediction model for real-time physical parameter data exceeds the warning threshold, a warning signal is issued.

[0017] To achieve the above objectives, a second embodiment of the present invention provides a comprehensive monitoring and early warning device for surrounding rock in a roadway, comprising the following modules:

[0018] A construction module is used to construct a sensor network for the surrounding rock of the roadway. The sensor network is composed of multiple types of sensors arranged at multiple characteristic monitoring locations in the roadway. The characteristic monitoring locations are determined by the geological structure of the roadway and the geological characteristics of the surrounding rock. The multiple types of sensors include displacement sensors, stress sensors, acceleration sensors, temperature sensors, and humidity sensors.

[0019] an acquisition module, configured to acquire physical parameter data of the surrounding rock at each characteristic monitoring position in real time through the sensor network, wherein the physical parameter data includes displacement change data of the surrounding rock acquired by the displacement sensor, stress state data of the surrounding rock acquired by the stress sensor, vibration characteristic data of the surrounding rock acquired by the acceleration sensor, and environmental parameters in the tunnel acquired by the temperature sensor and the humidity sensor;

[0020] A transmission module, used to transmit the displacement change data, stress state data, vibration characteristic data, and environmental parameters to a preset monitoring center via a network;

[0021] A processing module, configured to filter, denoise, and remove outliers from the displacement change data, stress state data, vibration characteristic data, and environmental parameters through the monitoring center to obtain standard multi-source data;

[0022] An analysis module, configured to comprehensively analyze the multi-source data using a method combining Bayesian estimation and fuzzy logic, and extract a stability score of the surrounding rock stability;

[0023] The training module is used to take physical parameter data as input and the stability score of the surrounding rock stability as output, and to train based on the machine learning algorithm to establish a prediction model for surrounding rock stability;

[0024] The early warning module is used to set the early warning threshold. When the real-time stability score predicted by the prediction model for real-time physical parameter data exceeds the early warning threshold, an early warning signal is issued.

[0025] To achieve the above-mentioned purpose, the third aspect embodiment of the present invention proposes an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method described in the first aspect.

[0026] In order to achieve the above-mentioned objectives, an embodiment of the fourth aspect of the present invention proposes a non-transitory computer-readable storage medium storing computer instructions, where the computer instructions are used to enable the computer to execute the method described in the first aspect.

[0027] The comprehensive monitoring and early warning method, device, electronic device and storage medium of the tunnel surrounding rock of the embodiment of the present invention constructs a sensor network of the tunnel surrounding rock; collects physical parameter data of the surrounding rock in real time through the sensor network; transmits the collected physical parameter data to the monitoring center through the network, and performs filtering, denoising and outlier removal to obtain standard multi-source data; uses a method combining Bayesian estimation and fuzzy logic to comprehensively analyze the multi-source data and extract the stability score of the surrounding rock stability; trains the relationship between the physical parameter data and the stability score, establishes a prediction model for the surrounding rock stability, and issues an early warning when the predicted real-time stability score exceeds the early warning threshold. Thus, through the constructed prediction model for the surrounding rock stability, comprehensive, real-time monitoring and high-precision early warning of the tunnel surrounding rock are achieved, and the instability of the surrounding rock is discovered and handled in a timely manner, avoiding the occurrence of safety accidents.

[0028] Additional aspects and advantages of the present invention will be set forth in part in the description which follows and, in part, will be obvious from the description which follows, or may be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which:

[0030] Figure 1 A schematic flow chart of a method for comprehensive monitoring and early warning of surrounding rock in a roadway provided by an embodiment of the present invention;

[0031] Figure 2 This is a structural schematic diagram of a tunnel surrounding rock comprehensive monitoring and early warning device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0032] The following describes embodiments of the present invention in detail, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and are not to be construed as limiting the present invention.

[0033] It should be noted that the acquisition, storage, use, and processing of data in the technical solution of the present invention comply with the relevant provisions of relevant laws and regulations.

[0034] The following describes a method and device for comprehensive monitoring and early warning of surrounding rock in a roadway according to an embodiment of the present invention with reference to the accompanying drawings.

[0035] Figure 1 A schematic flow chart of a method for comprehensive monitoring and early warning of surrounding rock in a roadway provided by an embodiment of the present invention.

[0036] like Figure 1 As shown, the method includes the following steps:

[0037] Step 101, constructing a sensor network for the tunnel surrounding rock, wherein the sensor network is composed of multiple types of sensors arranged at multiple characteristic monitoring positions in the tunnel, the characteristic monitoring positions are determined by the geological structure of the tunnel and the geological characteristics of the surrounding rock, and the multiple types of sensors include displacement sensors, stress sensors, acceleration sensors, temperature sensors and humidity sensors.

[0038] In some embodiments, various types of sensors in the sensor network are arranged in a grid-like manner, and the spacing d between various types of sensors satisfies:

[0039] d≤d max

[0040] Among them, d max is the maximum allowable sensor spacing based on the spatial resolution of comprehensive monitoring of tunnel surrounding rock; the number of sensors N is determined by the area S of the tunnel surrounding rock monitoring area and the layout density D, N = D × S.

[0041] Specifically, according to the principle of grid layout, the spacing between various sensor types can be set to 10 meters to ensure the spatial resolution of tunnel surrounding rock monitoring and to promptly detect local anomalies in the surrounding rock. In key areas with complex geological structures, concentrated stresses, and historical geological disasters, the sensor layout density is appropriately increased to strengthen monitoring efforts in these areas. The required number of sensors is calculated based on the area and layout density of the tunnel surrounding rock monitoring area. For example, a total of 100 monitoring points are arranged, each equipped with the five types of sensors mentioned above, totaling 500 sensors, forming a high-density, high-reliability sensor network.

[0042] Step 102: The physical parameter data of the surrounding rock at each characteristic monitoring position are collected in real time through the sensor network, wherein the physical parameter data include displacement change data of the surrounding rock collected by the displacement sensor, stress state data of the surrounding rock collected by the stress sensor, vibration characteristic data of the surrounding rock collected by the acceleration sensor, and environmental parameters in the tunnel collected by the temperature sensor and the humidity sensor.

[0043] In some embodiments, when the sensor network collects physical parameter data of the surrounding rock at each characteristic monitoring position in real time, the sampling frequency f of each sensor in the sensor network is i and the data format is set as follows:

[0044] Sampling frequency:

[0045] The sampling frequency f1 of the displacement sensor and stress sensor satisfies:

[0046] f1≥1 time / minute

[0047] The sampling frequency f2 of the acceleration sensor satisfies:

[0048] f2≥100 times / second

[0049] The sampling frequency f3 of the temperature and humidity sensor satisfies:

[0050] f3≥1 time / hour

[0051] Data format: The data collected by each sensor includes sensor ID, timestamp t, measurement value x i (t) and units. Ensure the traceability and consistency of data to facilitate subsequent data processing and analysis.

[0052] Optionally, displacement sensors are used to monitor the displacement change data of the surrounding rock to capture the deformation information of the tunnel; stress sensors are used to measure the stress state data of the surrounding rock to reflect the stress condition of the surrounding rock; acceleration sensors are used to monitor the vibration characteristic data of the surrounding rock to capture possible geological disaster signals; temperature and humidity sensors are used to monitor the environmental parameters in the tunnel to evaluate the impact of the environment on the stability of the surrounding rock.

[0053] Furthermore, after the sensor network is deployed, a high-precision data acquisition module collects real-time data on the physical parameters of the surrounding rock from various sensor types, ensuring data accuracy and timeliness. Sampling frequencies are set accordingly based on the characteristics and monitoring requirements of each sensor type. The displacement and stress sensors are sampled once per minute, enabling timely capture of deformation and stress changes in the roadway surrounding rock, reflecting its steady-state trends. The acceleration sensors, which need to capture transient vibration signals (vibration characteristic data) of the roadway surrounding rock, are sampled 100 times per second. This allows them to capture high-frequency vibration signals and promptly identify potential geological disaster precursors. The temperature and humidity sensors are sampled once per hour to meet environmental parameter monitoring requirements and assess the long-term impact of environmental factors on roadway surrounding rock stability. The data acquisition module boasts high precision, low power consumption, and strong anti-interference capabilities, enabling stable operation under the complex environmental conditions of the mine, ensuring continuous and reliable data acquisition.

[0054] Step 103: Transmit the displacement change data, stress state data, vibration characteristic data, and environmental parameters to a preset monitoring center via a network.

[0055] In some embodiments, when the displacement change data, stress state data, vibration characteristic data, and environmental parameters are transmitted to a preset monitoring center via a network, the network protocol uses a combination of industrial Ethernet and wireless mesh network, and data encryption, error correction coding, and data integrity verification are performed during the network transmission process, wherein:

[0056] Data encryption uses the Advanced Encryption Standard for data encryption. The data encryption formula is:

[0057] C=AES K (M)

[0058] Among them, C is the ciphertext, M is the plaintext data, and K is the key;

[0059] The error correction coding uses forward error correction code to encode the shift change data, stress state data, vibration characteristic data, and environmental parameters;

[0060] The data integrity check uses a cyclic redundancy check code, and the check polynomial is G(x).

[0061] Optionally, collected displacement change data, stress state data, vibration characteristic data, and environmental parameters are transmitted to a ground monitoring center via a network, using a combination of industrial Ethernet and wireless mesh networks to ensure the security, reliability, and real-time nature of data transmission. Industrial Ethernet lines are laid within the tunnels to achieve high-bandwidth, low-latency data transmission, meeting the transmission needs of large-scale data volumes. In areas where wired networks are unavailable, wireless mesh networks are used to achieve flexible network coverage and self-healing capabilities, ensuring network connectivity and stability. To prevent data theft and tampering during transmission, data is encrypted using the Advanced Encryption Standard (AES-256) to ensure data security and confidentiality. Cyclic Redundancy Check (CRC) technology is used to verify data integrity and prevent errors during transmission. Furthermore, forward error correction (FEC) technologies, such as Reed-Solomon codes, are used to encode data to improve its anti-interference capabilities during transmission and enhance data transmission reliability. During network transmission, the data transmission protocol adopts the industry standard Transmission Control Protocol / Internet Protocol (TCP / IP) protocol to ensure the compatibility and standardization of data transmission.

[0062] Step 104 : The monitoring center filters, removes noise, and eliminates outliers on the displacement change data, stress state data, vibration characteristic data, and environmental parameters to obtain standard multi-source data.

[0063] In some embodiments, the monitoring center filters, denoises, and removes outliers on displacement change data, stress state data, vibration characteristic data, and environmental parameters to obtain standard multi-source data including:

[0064] By performing a continuous wavelet transform on the signal x(t) corresponding to the vibration characteristic data, selecting a preset wavelet basis function and decomposition layer number J, filtering out high-frequency noise, and obtaining the first denoised multi-source signal, the continuous wavelet transform formula is:

[0065]

[0066] Among them, ψ(t) is the wavelet basis function, a is the scale parameter, b is the translation parameter, ψ * (t) is the conjugate of ψ(t);

[0067] The displacement change data and the stress state data are filtered by an adaptive filter to obtain a filtered second multi-source signal, wherein the weight update formula of the adaptive filter is:

[0068] w(n+1)=w(n)+μe(n)×x(n)

[0069] Where μ is the step size factor, e(n) = d(n) - y(n) is the error, d(n) is the desired signal, and y(n) = w T (n)×x(n) is the filter output, x(n) is the first multi-source signal, and w(n) is the weight vector;

[0070] The first multi-source signal, the second multi-source signal, and the environmental parameters are subjected to outlier elimination using the three-sigma criterion or the box plot method to obtain a third multi-source signal, which is used as the standard multi-source data, wherein:

[0071] When the first multi-source signal, the second multi-source signal, and the environmental parameters are subjected to outlier elimination using the three-sigma criterion, when the second multi-source signal deviates from the mean μ by more than three times the standard deviation σ, it is determined to be an outlier and meets the following conditions:

[0072] |x i -μ|>3σ

[0073] When the first multi-source signal, the second multi-source signal and the environmental parameters are subjected to outlier elimination by the box plot method, the quartiles Q1 and Q3 are calculated, and the outlier range is:

[0074] x i <=Q1-1.5×IQRx i and x i >Q3+1.5×IQR

[0075] Among them, IQR=Q3-Q1.

[0076] Furthermore, at the monitoring center, the received sensor data undergoes preprocessing, including filtering, denoising, and outlier removal, to obtain standardized multi-source data, improving data quality and laying the foundation for subsequent comprehensive data analysis. For acceleration signals, which may be affected by mechanical vibration and electromagnetic interference noise, a continuous wavelet transform (CWT) filtering method is used to effectively filter out high-frequency noise, retaining the useful vibration signal (the first multi-source signal), and extracting possible geological disaster precursors. For displacement change data and stress state data, which may be affected by environmental factors and measurement errors, an adaptive filter is used to automatically adjust the filtering parameters based on the signal characteristics, improving the filtering effect and accurately reflecting the deformation and stress state of the tunnel surrounding rock. For outlier removal, the three-sigma criterion and box plot method are used to calculate the mean, standard deviation, and quartiles of the data. Data points that deviate from the mean by more than three standard deviations and fall outside the upper and lower limits of the box plot are considered outliers and removed to prevent abnormal data from influencing the analysis results. Furthermore, the data is normalized to eliminate the influence of different sensor data dimensions and ensure that the data are on the same scale.

[0077] Step 105 , using a method combining Bayesian estimation and fuzzy logic, comprehensively analyzes the multi-source data and extracts a stability score of the surrounding rock stability.

[0078] In some embodiments, the posterior probability calculation formula of the Bayesian estimation is:

[0079]

[0080] Where θ is the parameter to be estimated, D is the observed data, P(θ) is the prior probability, P(θ|D) is the likelihood function, and P(D) is the marginal probability.

[0081] Fuzzy logic consists of membership function definition, building a fuzzy rule base, fuzzy reasoning and defuzzification process. The membership function definition is to establish a fuzzy membership function μ(x) for each parameter of multi-source data.

[0082]

[0083] Among them, a and b are parameter thresholds;

[0084] The fuzzy rule base is constructed by formulating fuzzy rules. The fuzzy rules are as follows: if the displacement change data corresponds to a “large” displacement and the stress state data corresponds to a “high” stress, then the surrounding rock stability is “low”;

[0085] Fuzzy reasoning uses Mamdani reasoning method to calculate the output of surrounding rock stability evaluation;

[0086] The defuzzification process uses the centroid method to calculate the final output value y, the formula is:

[0087]

[0088] Among them, μ(y ' ) is the membership function of the output variable.

[0089] Specifically, multi-source data is comprehensively analyzed to extract a stability score for surrounding rock stability, comprehensively assessing the stability of the roadway surrounding rock. Using Bayesian estimation, prior knowledge and observed data are combined to provide a probabilistic estimate of the roadway surrounding rock stability, improving the accuracy and reliability of the estimate. Fuzzy membership functions are established for each multi-source data set. Based on expert experience and historical data, displacement change data, stress state data, and vibration characteristic data are classified into different fuzzy sets, such as "small," "medium," "large," and "low," "medium," and "high." A fuzzy rule base is developed, such as "If displacement is large and stress is high, stability is low." Through fuzzy reasoning and combining the membership degrees of each multi-source data set, a comprehensive evaluation of the roadway surrounding rock stability is obtained, such as "stable," "caution needed," and "unstable." During the data fusion process, different types of sensor data are combined according to specific weights. The weights are determined based on the sensor's reliability, sensitivity, and impact on surrounding rock stability. For example, displacement change data and stress state data are assigned higher weights, while environmental parameters such as temperature and humidity are assigned appropriate weights. This comprehensively reflects the state of the roadway surrounding rock and improves the accuracy of the assessment results.

[0090] Step 106 , using the physical parameter data as input and the stability score of the surrounding rock stability as output, training is performed based on a machine learning algorithm to establish a prediction model for surrounding rock stability.

[0091] In some embodiments, the machine learning algorithm uses a deep learning model that combines a long short-term memory network and a convolutional neural network for prediction, wherein:

[0092] The state update formula of the long short-term memory network is:

[0093]

[0094] Among them, σ is the Sigmoid function, tanh is the hyperbolic tangent function, ⊙ is the element product, h t is the hidden state C t is the cell state, x t is the input, W and b are the weight matrix and bias vector;

[0095] The convolution calculation formula of the convolutional neural network is:

[0096]

[0097] Among them, x is the input feature map, w is the convolution kernel, y is the output feature map, M and N are the convolution kernel sizes;

[0098] When training a deep learning model, the loss function uses mean square error, and the formula is:

[0099]

[0100] Among them, y i is the true value, is the predicted value, N is the number of samples;

[0101] The optimization algorithm of the deep learning model uses the gradient descent algorithm and the Adam optimizer to adjust the model parameters to minimize the loss function. The methods to prevent overfitting of the deep learning model include Dropout and early stopping to prevent overfitting.

[0102] Furthermore, a Long Short-Term Memory (LSTM) network was used to capture time series features, analyze temporal trends in the physical parameter data of the roadway surrounding rock, and predict future changes. A Convolutional Neural Network (CNN) was used to extract spatial features, analyze the spatial correlation of data from different monitoring locations in the sensor network, and detect local anomalies. During the prediction model training process, historical physical parameter data was used to adjust and optimize the model parameters. Overfitting prevention strategies, such as Dropout and early stopping, were employed to ensure the model's generalization ability. During the model validation process, a subset of physical parameter data was used to evaluate the model's predictive performance, such as mean squared error and accuracy, to ensure its reliability. The trained prediction model was applied to real-time physical parameter data to predict the future stability of the roadway surrounding rock, promptly identifying potential safety hazards and providing early warning information.

[0103] Step 107 , setting a warning threshold. When the real-time stability score predicted by the prediction model for the real-time physical parameter data exceeds the warning threshold, a warning signal is issued to issue a warning.

[0104] In some embodiments, a warning threshold is set. When the real-time stability score predicted by the prediction model for the real-time physical parameter data exceeds the warning threshold, a warning signal is issued to issue a warning, including:

[0105] The warning thresholds are set to include the first-level warning threshold T1 = 0.6, the second-level warning threshold T2 = 0.7 and the third-level warning threshold T3 = 0.8. Each level of warning threshold corresponds to a different warning level L i ;

[0106] When the real-time stability score S exceeds the warning threshold, the risk index R of the tunnel surrounding rock is calculated. i , when i=1,2,3, the corresponding warning level L is triggered i , issue early warning signals for early warning;

[0107] The calculation formula of risk index R is:

[0108]

[0109] The calculation formula of stability score S is: S=w1S monitor +w2S predict ,S monitor Score the monitoring of real-time physical parameter data. predict is the score of the prediction model, w1 and w2 are weight coefficients, satisfying w1+w2=1; S min and S max are the minimum and maximum stability scores, respectively.

[0110] Specifically, for example, a risk index of 0.6 or above triggers a first-level warning (attention), prompting relevant personnel to strengthen monitoring and inspections; a risk index of 0.7 or above triggers a second-level warning (warning), prompting relevant personnel to take preventive measures, such as strengthening support and restricting entry; a risk index of 0.8 or above triggers a third-level warning (danger), prompting immediate emergency measures, such as evacuation of personnel and cessation of operations.

[0111] In addition, the early warning rule base {R j} Perform secondary discrimination on real-time physical parameter data and real-time stability score. Secondary discrimination is done by fuzzy reasoning. In fuzzy reasoning, the rule matching degree {μ j} is calculated as:

[0112]

[0113] Among them, μ Aji (x i ) is the variable x i In the warning rule base {R j}Inside rule R j The membership degree in , the comprehensive matching degree μ is calculated as:

[0114] μ=max{μ j}

[0115] When the comprehensive matching μ ≥ μ threshold When the warning signal is confirmed to be effective, μ threshold is the preset matching threshold.

[0116] Therefore, by analyzing abnormal situations according to the established early warning rule base, the effectiveness of the early warning can be verified, the false alarm rate can be reduced, and the accuracy of the early warning can be improved.

[0117] Furthermore, the warning level L corresponding to the warning signal can be i and tunnel surrounding rock types to generate multiple emergency plans and disposal suggestions, and push them to management personnel, including emergency plans Including personnel evacuation plan, support and reinforcement measures, monitoring and strengthening plan, in the emergency plan In the case of a personnel evacuation plan, the formula for calculating the evacuation time is:

[0118]

[0119] Among them, D exit is the distance to the nearest tunnel exit, v avg is the average travel speed of personnel;

[0120] In the emergency plan In the case of support reinforcement measures, calculate the anchor support force F anchor ,

[0121] F anchor =n×A s ×f y

[0122] Among them, n is the number of anchor rods corresponding to the support, A s is the cross-sectional area of ​​the anchor rod, f y is the yield strength of the anchor rod.

[0123] Specifically, the warning level corresponding to the warning signal will be notified to the management personnel through multiple channels and emergency plans and disposal suggestions will be provided. Among them, multi-channel notifications include but are not limited to on-site sound and light alarms (setting up sound and light alarms, and when the warning is triggered, sending sound and light alarms to mobile terminals), real-time push of warning information through mobile phone applications (content includes warning level, time, location, risk index R, disposal suggestions), email and SMS notifications, and automatic generation of warning reports, which will be sent to the management personnel's mailbox and mobile phone.

[0124] Optionally, real-time stability score curves, historical stability score trend charts, and predicted stability score curves can be provided at the monitoring center and on mobile terminals to visualize the state of the tunnel surrounding rock and assist management in decision-making. Using time series analysis methods, trend predictions can be made on displacement change data and stress state data to assess the changing trends in the tunnel surrounding rock stability and provide a scientific basis for decision-making. This method is not only applicable to surrounding rock monitoring and early warning in mine tunnels, but can also be extended to tunnel construction, underground space development, and other underground engineering fields.

[0125] The comprehensive monitoring and early warning method for tunnel surrounding rock of the embodiment of the present invention constructs a sensor network for tunnel surrounding rock; collects physical parameter data of surrounding rock in real time through the sensor network; transmits the collected physical parameter data to the monitoring center through the network, and performs filtering, denoising and outlier removal to obtain standard multi-source data; uses a method combining Bayesian estimation and fuzzy logic to comprehensively analyze the multi-source data and extract the stability score of the surrounding rock stability; trains the relationship between the physical parameter data and the stability score, establishes a prediction model for surrounding rock stability, and issues an early warning when the predicted real-time stability score exceeds the early warning threshold. Thus, through the constructed prediction model for surrounding rock stability, comprehensive, real-time monitoring and high-precision early warning of tunnel surrounding rock are achieved, and unstable conditions of surrounding rock are discovered and handled in a timely manner, thus avoiding the occurrence of safety accidents.

[0126] In order to implement the above embodiment, the present invention also proposes a comprehensive monitoring and early warning device for tunnel surrounding rocks.

[0127] Figure 2 A schematic structural diagram of a tunnel surrounding rock comprehensive monitoring and early warning device provided in an embodiment of the present invention.

[0128] like Figure 2 As shown, the tunnel surrounding rock comprehensive monitoring and early warning device 20 includes: a construction module 21, a collection module 22, a transmission module 23, a processing module 24, an analysis module 25, a training module 26, and an early warning module 27.

[0129] A construction module 21 is used to construct a sensor network for the surrounding rock of the roadway, wherein the sensor network is composed of multiple types of sensors arranged at multiple characteristic monitoring locations in the roadway, wherein the characteristic monitoring locations are determined by the geological structure of the roadway and the geological characteristics of the surrounding rock, and the multiple types of sensors include displacement sensors, stress sensors, acceleration sensors, temperature sensors, and humidity sensors;

[0130] an acquisition module 22 for collecting physical parameter data of the surrounding rock at each characteristic monitoring position in real time through the sensor network, wherein the physical parameter data includes displacement change data of the surrounding rock collected by the displacement sensor, stress state data of the surrounding rock collected by the stress sensor, vibration characteristic data of the surrounding rock collected by the acceleration sensor, and environmental parameters in the tunnel collected by the temperature sensor and the humidity sensor;

[0131] The transmission module 23 is used to transmit the displacement change data, stress state data, vibration characteristic data, and environmental parameters to a preset monitoring center via a network;

[0132] The processing module 24 is used to filter, denoise and remove outliers on the displacement change data, stress state data, vibration characteristic data and environmental parameters through the monitoring center to obtain standard multi-source data;

[0133] An analysis module 25 is configured to perform a comprehensive analysis on the multi-source data using a method combining Bayesian estimation and fuzzy logic to extract a stability score of the surrounding rock stability;

[0134] A training module 26 is configured to take the physical parameter data as input and the stability score of the surrounding rock stability as output, perform training based on a machine learning algorithm, and establish a prediction model for the surrounding rock stability;

[0135] The early warning module 27 is used to set an early warning threshold. When the real-time stability score predicted by the prediction model for the real-time physical parameter data exceeds the early warning threshold, an early warning signal is issued to issue an early warning.

[0136] The comprehensive monitoring and early warning device for tunnel surrounding rock of the embodiment of the present invention constructs a sensor network for tunnel surrounding rock; collects physical parameter data of the surrounding rock in real time through the sensor network; transmits the collected physical parameter data to the monitoring center through the network, and performs filtering, denoising, and outlier removal to obtain standard multi-source data; utilizes a method combining Bayesian estimation and fuzzy logic to comprehensively analyze the multi-source data and extract the stability score of the surrounding rock stability; trains the relationship between the physical parameter data and the stability score, establishes a prediction model for the surrounding rock stability, and issues an early warning when the predicted real-time stability score exceeds the early warning threshold. Thus, through the constructed prediction model for surrounding rock stability, comprehensive, real-time monitoring and high-precision early warning of tunnel surrounding rock are achieved, and unstable conditions of the surrounding rock are discovered and handled in a timely manner, thus avoiding the occurrence of safety accidents.

[0137] In order to implement the above embodiment, the present invention further provides an electronic device, including:

[0138] at least one processor; and

[0139] a memory communicatively connected to the at least one processor; wherein,

[0140] The memory stores instructions that can be executed by the at least one processor. The instructions are executed by the at least one processor to enable the at least one processor to perform the aforementioned method.

[0141] In order to implement the above embodiment, the present invention further proposes a non-transitory computer-readable storage medium storing computer instructions, where the computer instructions are used to enable the computer to execute the above method.

[0142] In the description of this specification, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0143] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present invention, "plurality" means at least two, such as two, three, etc., unless otherwise specifically defined.

[0144] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a custom logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention pertain.

[0145] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and programmable read-only memory (EPROM or flash memory), fiber optic devices, and a portable compact disc read-only memory (CDROM). Furthermore, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing it in another suitable manner if necessary, and then storing it in a computer memory.

[0146] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0147] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0148] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing module, or each unit may exist physically separately, or two or more units may be integrated into a single module. The aforementioned integrated modules may be implemented in the form of hardware or in the form of software functional modules. If the integrated modules are implemented in the form of software functional modules and sold or used as independent products, they may also be stored in a computer-readable storage medium.

[0149] The storage medium mentioned above may be a read-only memory, a magnetic disk, or an optical disk, etc. Although the embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and are not to be construed as limiting the present invention. Persons skilled in the art may make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present invention.

Claims

1. A comprehensive monitoring and early warning method for tunnel surrounding rock, characterized in that: The following steps are involved: Construct a sensor network for the tunnel surrounding rock. The sensor network consists of multiple types of sensors deployed at multiple characteristic monitoring locations in the tunnel. The characteristic monitoring locations are determined by the geological structure of the tunnel and the geological characteristics of the surrounding rock. The multiple types of sensors include displacement sensors, stress sensors, acceleration sensors, temperature sensors, and humidity sensors. The sensor network collects physical parameter data of the surrounding rock at each characteristic monitoring position in real time, wherein the physical parameter data includes displacement change data of the surrounding rock collected by the displacement sensor, stress state data of the surrounding rock collected by the stress sensor, vibration characteristic data of the surrounding rock collected by the acceleration sensor, and environmental parameters in the tunnel collected by the temperature sensor and the humidity sensor; Transmitting the displacement change data, stress state data, vibration characteristic data, and environmental parameters to a preset monitoring center via a network; The monitoring center filters, removes noise and eliminates outliers on displacement change data, stress state data, vibration characteristic data and environmental parameters to obtain standard multi-source data; By combining Bayesian estimation and fuzzy logic, the multi-source data are comprehensively analyzed to extract the stability score of the surrounding rock stability; Taking physical parameter data as input and the stability score of surrounding rock stability as output, a prediction model of surrounding rock stability is established through training based on machine learning algorithm; Set a warning threshold. When the real-time stability score predicted by the prediction model for real-time physical parameter data exceeds the warning threshold, a warning signal is issued.

2. A tunnel surrounding rock comprehensive monitoring and early warning method according to claim 1, characterized in that: The various types of sensors in the sensor network are arranged in a grid-like manner, and the spacing d between the various types of sensors satisfies: d≤d max Among them, d max The maximum sensor spacing allowed is the spatial resolution based on comprehensive monitoring of tunnel surrounding rock; The number of sensors N is determined by the area S of the tunnel surrounding rock monitoring area and the layout density D, N = D × S.

3. A tunnel surrounding rock comprehensive monitoring and early warning method according to claim 1, characterized in that: When the sensor network collects the physical parameter data of the surrounding rock at each characteristic monitoring position in real time, the sampling frequency f of each sensor in the sensor network is i and the data format is set as follows: Sampling frequency: The sampling frequency f1 of the displacement sensor and stress sensor satisfies: f1≥1 time / minute The sampling frequency f2 of the acceleration sensor satisfies: f2≥100 times / second The sampling frequency f3 of the temperature and humidity sensor satisfies: f3≥1 time / hour Data format: The data collected by each sensor includes sensor ID, timestamp t, measurement value x i (t) and units.

4. A tunnel surrounding rock comprehensive monitoring and early warning method according to claim 1, characterized in that: When the displacement change data, stress state data, vibration characteristic data, and environmental parameters are transmitted to a preset monitoring center via a network, the network protocol adopts a combination of industrial Ethernet and wireless mesh network, and data encryption, error correction coding, and data integrity verification are performed during the network transmission process, wherein: Data encryption uses the Advanced Encryption Standard for data encryption. The data encryption formula is: C=AES K (M) Among them, C is the ciphertext, M is the plaintext data, and K is the key; The error correction coding uses forward error correction code to encode the shift change data, stress state data, vibration characteristic data, and environmental parameters; The data integrity check uses a cyclic redundancy check code, and the check polynomial is G(x).

5. A comprehensive monitoring and early warning method for tunnel surrounding rock according to claim 1, characterized in that: The monitoring center filters, removes noise, and eliminates outliers on displacement change data, stress state data, vibration characteristic data, and environmental parameters to obtain standard multi-source data including: By performing a continuous wavelet transform on the signal x(t) corresponding to the vibration characteristic data, selecting a preset wavelet basis function and decomposition layer number J, filtering out high-frequency noise, and obtaining the first denoised multi-source signal, the continuous wavelet transform formula is: Among them, ψ(t) is the wavelet basis function, a is the scale parameter, b is the translation parameter, ψ * (t) is the conjugate of ψ(t); The displacement change data and the stress state data are filtered by an adaptive filter to obtain a filtered second multi-source signal, wherein the weight update formula of the adaptive filter is: w(n+1)=w(n)+μe(n)×x(n) Where μ is the step size factor, e(n) = d(n) - y(n) is the error, d(n) is the desired signal, and y(n) = w T (n)×x(n) is the filter output, x(n) is the first multi-source signal, and w(n) is the weight vector; The first multi-source signal, the second multi-source signal, and the environmental parameters are subjected to outlier elimination using the three-sigma criterion or the box plot method to obtain a third multi-source signal, which is used as the standard multi-source data, wherein: When the first multi-source signal, the second multi-source signal, and the environmental parameters are subjected to outlier elimination using the three-sigma criterion, when the second multi-source signal deviates from the mean μ by more than three times the standard deviation σ, it is determined to be an outlier and meets the following conditions: |x i -μ|>3σ When the first multi-source signal, the second multi-source signal and the environmental parameters are subjected to outlier elimination by the box plot method, the quartiles Q1 and Q3 are calculated, and the outlier range is: x i <=Q1-1.5×IQRx i and x i >Q3+1.5×IQR Among them, IQR=Q3-Q1.

6. A tunnel surrounding rock comprehensive monitoring and early warning method according to claim 1, characterized in that: The formula for calculating the posterior probability of Bayesian estimation is: Where θ is the parameter to be estimated, D is the observed data, P(θ) is the prior probability, P(θ|D) is the likelihood function, and P(D) is the marginal probability. Fuzzy logic consists of membership function definition, building a fuzzy rule base, fuzzy reasoning and defuzzification process. The membership function definition is to establish a fuzzy membership function μ(x) for each parameter of multi-source data. Among them, a and b are parameter thresholds; The fuzzy rule base is constructed by formulating fuzzy rules. The fuzzy rules are as follows: if the displacement change data corresponds to a "large" displacement and the stress state data corresponds to a "high" stress, then the surrounding rock stability is "low"; Fuzzy reasoning uses Mamdani reasoning method to calculate the output of surrounding rock stability evaluation; The defuzzification process uses the centroid method to calculate the final output value y, the formula is: Among them, μ(y') is the membership function of the output variable.

7. A tunnel surrounding rock comprehensive monitoring and early warning method according to claim 1, characterized in that: The machine learning algorithm uses a deep learning model that combines long short-term memory networks and convolutional neural networks for prediction, where: The state update formula of the long short-term memory network is: Among them, σ is the Sigmoid function, tanh is the hyperbolic tangent function, ⊙ is the element product, h t is the hidden state C t is the cell state, x t is the input, W and b are the weight matrix and bias vector; The convolution calculation formula of the convolutional neural network is: Among them, x is the input feature map, w is the convolution kernel, y is the output feature map, M and N are the convolution kernel sizes; When the deep learning model is trained, the loss function uses the mean square error, and the formula is: Among them, y i is the true value, is the predicted value, N is the number of samples; The optimization algorithm of the deep learning model adopts the gradient descent algorithm and the Adam optimizer, and the method for preventing overfitting of the deep learning model includes regularization Dropout and early stopping method.

8. A tunnel surrounding rock comprehensive monitoring and early warning method according to claim 1, characterized in that: The setting of the warning threshold, when the real-time stability score predicted by the prediction model for the real-time physical parameter data exceeds the warning threshold, issuing a warning signal for warning, includes: The set warning thresholds include the first-level warning threshold T1=0.6, the second-level warning threshold T2=0.7 and the third-level warning threshold T3=0.

8. Each level of warning threshold corresponds to a different warning level L i ; When the real-time stability score S exceeds the warning threshold, the risk index R of the tunnel surrounding rock is calculated. i , when i=1,2,3, the corresponding warning level L is triggered i , issue early warning signals for early warning; The calculation formula of risk index R is: The calculation formula of stability score S is: S=w1S monitor +w2S predict ,S monitor Score the monitoring of real-time physical parameter data. predict is the score of the prediction model, w1 and w2 are weight coefficients, satisfying w1+w2=1; S min and S max are the minimum and maximum stability scores, respectively.

9. A tunnel surrounding rock comprehensive monitoring and early warning method according to claim 8, characterized in that: The method further comprises: The early warning rule base {R j } Perform secondary discrimination on real-time physical parameter data and real-time stability score. Secondary discrimination is done by fuzzy reasoning. In fuzzy reasoning, the rule matching degree {μ j } is calculated as: Among them, μ Aji (x i ) is the variable x i In the warning rule base {R j }Inside rule R j The membership degree in , the comprehensive matching degree μ is calculated as: μ=max{μ j } When the comprehensive matching μ ≥ μ threshold When the warning signal is confirmed to be effective, μ threshold is the preset matching threshold.

10. A tunnel surrounding rock comprehensive monitoring and early warning method according to claim 8, characterized in that: The method further comprises: According to the warning level L corresponding to the warning signal i and tunnel surrounding rock types to generate multiple emergency plans and disposal suggestions, and push them to management personnel, including emergency plans Including personnel evacuation plan, support and reinforcement measures, monitoring and strengthening plan, in the emergency plan In the case of a personnel evacuation plan, the formula for calculating the evacuation time is: Among them, D exit is the distance to the nearest tunnel exit, v avg is the average travel speed of personnel; In the emergency plan In the case of support reinforcement measures, calculate the anchor support force F anchor , F anchor =n×A s ×f y Among them, n is the number of anchor rods corresponding to the support, A s is the cross-sectional area of ​​the anchor rod, f y is the yield strength of the anchor rod.

11. A comprehensive monitoring and early warning device for tunnel surrounding rock, characterized in that: Includes the following modules: A construction module is used to construct a sensor network for the surrounding rock of the roadway. The sensor network is composed of multiple types of sensors arranged at multiple characteristic monitoring locations in the roadway. The characteristic monitoring locations are determined by the geological structure of the roadway and the geological characteristics of the surrounding rock. The multiple types of sensors include displacement sensors, stress sensors, acceleration sensors, temperature sensors, and humidity sensors. an acquisition module, configured to acquire physical parameter data of the surrounding rock at each characteristic monitoring position in real time through the sensor network, wherein the physical parameter data includes displacement change data of the surrounding rock acquired by the displacement sensor, stress state data of the surrounding rock acquired by the stress sensor, vibration characteristic data of the surrounding rock acquired by the acceleration sensor, and environmental parameters in the tunnel acquired by the temperature sensor and the humidity sensor; A transmission module, used to transmit the displacement change data, stress state data, vibration characteristic data, and environmental parameters to a preset monitoring center via a network; A processing module, configured to filter, denoise, and remove outliers from the displacement change data, stress state data, vibration characteristic data, and environmental parameters through the monitoring center to obtain standard multi-source data; An analysis module, configured to comprehensively analyze the multi-source data using a method combining Bayesian estimation and fuzzy logic, and extract a stability score of the surrounding rock stability; The training module is used to take physical parameter data as input and the stability score of the surrounding rock stability as output, and to train based on the machine learning algorithm to establish a prediction model for surrounding rock stability; The early warning module is used to set the early warning threshold. When the real-time stability score predicted by the prediction model for real-time physical parameter data exceeds the early warning threshold, an early warning signal is issued.

12. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 10.

13. A non-transitory computer-readable storage medium storing computer instructions, characterized in that: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-10.

Images