Monitoring and early warning processing system and method for downhole construction site

Abstract

The application relates to the technical field of underground roadway monitoring, in particular to a monitoring and early warning processing system and method for an underground construction site, which comprises a service end and a management end; the service end comprises a data acquisition module, an identification module, a calculation module, a prediction module and an early warning module; the prediction module is used for calling a first monitoring and early warning prediction model corresponding to an abnormal sensor and a normal sensor after completing the calculation of sensor data of all abnormal sensors, inputting sensor data corresponding to the abnormal sensor and sensor data corresponding to the normal sensor into the corresponding first monitoring and early warning prediction model as input data, and outputting a first prediction result corresponding to the underground roadway at the current time; the early warning module is used for judging whether the underground roadway at the current time is dangerous according to the first prediction result corresponding to the underground roadway at the current time, and sending first early warning information to the management end if the underground roadway at the current time is dangerous.

Description

Technical Field

[0001] This invention relates to the field of underground roadway monitoring technology, specifically to a monitoring and early warning processing system and method for underground construction sites. Background Technology

[0002] Underground construction sites present extremely complex and variable environments, with a variety of potential safety risks, including but not limited to gas explosions, roof collapses, fires, and leaks of hazardous gases. These risks not only threaten the lives of construction workers but can also lead to production disruptions and significant economic losses. Therefore, real-time monitoring and early warning systems are crucial for ensuring the safety of construction workers and the smooth operation of production activities.

[0003] Traditional downhole monitoring and early warning systems primarily rely on direct sensor data acquisition and simple data analysis methods. However, this type of system has several significant limitations:

[0004] Each sensor operates independently, lacking comprehensive analysis of data from multiple sensors. When a sensor node experiences performance degradation, signal distortion, or complete failure, the key parameters it is responsible for monitoring cannot be effectively collected and transmitted, thus affecting the data integrity and accuracy of the entire monitoring network. Furthermore, the failure of a single sensor not only leads to localized information loss but may also trigger a chain reaction, interfering with the assessment of the operational status of other normally functioning units, and even causing a significant decrease in the reliability of the entire downhole monitoring system.

[0005] Therefore, there is an urgent need for a monitoring and early warning processing system and method for downhole construction sites, which can solve the problem that the reliability of downhole site monitoring is greatly reduced due to the loss of local information caused by the failure of some sensors in the existing technology, and greatly improve the reliability of downhole site monitoring when some sensors fail. Summary of the Invention

[0006] One of the objectives of this invention is to provide a monitoring and early warning processing system and method for downhole construction sites, which can solve the problem in the prior art where the failure of some sensors leads to the loss of local information, thereby significantly reducing the reliability of the entire downhole site monitoring. This invention greatly improves the reliability of downhole site monitoring when some sensors fail.

[0007] To achieve the above objectives, a monitoring and early warning processing system for downhole construction sites is provided, including a server and a management terminal;

[0008] The server includes:

[0009] The data acquisition module is used to collect sensor data from each sensor at each sensor point in a certain monitoring group within the underground roadway at the current moment, based on a pre-defined sensor point setting scheme.

[0010] The identification module is used to determine whether there is any abnormality in the sensor data of the corresponding sensor at each sensor location at the current time, based on the collected sensor data and the number of sensor points in the corresponding sensor location setting scheme. If so, the corresponding sensor is identified as an abnormal sensor; otherwise, the corresponding sensor is identified as a normal sensor. This process continues until all sensors are identified, forming their respective sets of abnormal and normal sensors.

[0011] The calculation module is used to select the data correlation set corresponding to each abnormal sensor in the abnormal sensor set based on the sensor data of the normal sensor set corresponding to the normal sensor set at the current time and a preset data correlation selection strategy. Based on the data correlation set corresponding to each abnormal sensor and the sensor data of the normal sensor corresponding to the data correlation, the module calculates the sensor data corresponding to the abnormal sensor.

[0012] The prediction module is used to retrieve the first monitoring and early warning models corresponding to the abnormal sensors and normal sensors after completing the calculation of the sensor data of all abnormal sensors, and input the sensor data corresponding to the abnormal sensors and the sensor data corresponding to the normal sensors as input data into the corresponding first monitoring and early warning prediction model, and output the first prediction result corresponding to the underground roadway at the current moment.

[0013] The early warning module is used to determine whether there is any danger in the underground roadway at the current moment based on the first prediction result corresponding to the current moment. If so, it sends the first early warning information to the management terminal.

[0014] The technical principle and effects of this solution are as follows: First, sensor data is collected from each sensor location corresponding to a monitoring group in the underground roadway at the current moment. Then, based on the collected sensor data and the number of sensor locations in the corresponding sensor location setup, it is sequentially determined whether there are any anomalies in the sensor data at each sensor location at the current moment. This step distinguishes between abnormal and normal sensor readings, automatically detecting and differentiating them. This process is crucial for the early detection of potential safety hazards, as abnormal data may be a precursor to certain potential problems.

[0015] Then, based on the sensor data of the normal sensors corresponding to the normal sensor set at the current moment, and based on the preset data correlation selection strategy, the data correlation set corresponding to each abnormal sensor in the abnormal sensor set is selected. Based on each data correlation set corresponding to each abnormal sensor, and the sensor data of the normal sensor corresponding to the data correlation set, the sensor data corresponding to the abnormal sensor is calculated. By obtaining the data correlation set corresponding to each abnormal sensor, the sensor data of the abnormal sensor is accurately calculated and supplemented, thereby improving the accuracy of anomaly detection.

[0016] Then, after completing the calculation of sensor data from all abnormal sensors, the first monitoring and early warning model corresponding to the abnormal and normal sensors is retrieved, and the sensor data corresponding to the abnormal sensors and the sensor data corresponding to the normal sensors are input into the corresponding first monitoring and early warning prediction model. The first prediction result corresponding to the underground roadway at the current moment is output. Based on the first prediction result corresponding to the underground roadway at the current moment, it is determined whether there is danger in the underground roadway at the current moment. If so, the first early warning information is sent to the management terminal.

[0017] Compared to existing technologies where a single sensor node experiences performance degradation, signal distortion, or complete failure, rendering the key parameters it monitors ineffective for data acquisition and transmission, thus affecting the data integrity and accuracy of the entire monitoring network, this solution addresses the issue of partial sensor failure. It addresses the problem by leveraging the collaborative work of the identification and computation modules to accurately and reliably supplement and improve the sensor data from the malfunctioning sensors. This data, along with data from the normal sensors, is then input into the corresponding model, significantly improving the reliability of downhole monitoring and ensuring its accuracy even when some sensors malfunction. This solution solves the problem of significantly reduced reliability in downhole monitoring due to partial sensor failure, greatly improving the reliability of downhole monitoring even when some sensors fail.

[0018] Furthermore, the preset data correlation selection strategy includes the following steps:

[0019] S100. Based on the normal sensor set and the abnormal sensor set at the current moment, calculate the ratio of the number of sensors between the normal sensor set and the abnormal sensor set;

[0020] S200. If the sensor quantity ratio is greater than or equal to the preset quantity ratio, then based on the grouping situation corresponding to the previous moment, the normal sensors in the normal sensor set are grouped to form various sensor groups; the grouping situation includes the number of groups and the sensors corresponding to each group.

[0021] S300. If the number ratio of sensors is less than the preset number ratio, then according to each normal sensor in the normal sensor set and the number of groups corresponding to the previous moment, each normal sensor in the normal sensor set is randomly selected and combined to form a sensor group with the same number of groups as the previous moment.

[0022] S400. Based on each sensor in each sensor group, retrieve the data correlation degree between each abnormal sensor in the abnormal sensor set and all normal sensors in each sensor group.

[0023] S500: Obtain the basic roadway data of the corresponding monitoring group in the underground roadway and the installation location data of the monitoring group in the underground roadway, and dynamically adjust the correlation of the retrieved data based on the basic roadway data and the installation location data.

[0024] S500. Based on the data correlation degree between each abnormal sensor in the dynamically adjusted abnormal sensor set and all normal sensors in each sensor group, as well as the sensor data corresponding to all normal sensors in the sensor group, calculate the sensor data corresponding to each abnormal sensor in the abnormal sensor set.

[0025] Beneficial effects: In this scheme, by first calculating the ratio of the number of sensors between the normal sensor set and the abnormal sensor set, the proportion of abnormal sensors in the corresponding monitoring group can be known.

[0026] When the sensor quantity ratio is greater than or equal to a preset ratio, the normal sensors in the normal sensor set are grouped based on the grouping situation at the previous time step, forming sensor groups. The grouping situation includes the number of groups and the sensors corresponding to each group. This approach fully considers that when the sensor quantity ratio meets the preset ratio, the proportion of damaged sensors is relatively small. Therefore, when using the grouping from the previous time step, the subsequent calculation results are reliable and accurate. Grouping based on the previous time step maintains the continuity and stability of sensor grouping. This continuity helps reduce unnecessary interference caused by frequent grouping adjustments, making the system smoother and more efficient in data processing. Utilizing existing grouping information avoids the large amount of computational resources required for regrouping from scratch. This not only saves time and computational costs but also improves the system's response speed, especially in large-scale sensor networks, where this advantage is particularly significant. Inheriting historical grouping information ensures the consistency of data from sensors within the same group at different time points, facilitating time series analysis and trend prediction. This is highly beneficial for long-term monitoring and fault diagnosis.

[0027] If the number of sensors is less than a preset ratio, then based on the number of normal sensors in the normal sensor set and the number of groups corresponding to the previous time step, the normal sensors in the normal sensor set are randomly selected and combined to form sensor groups with the same number of groups as the previous time step. While ensuring the number of groups remains the same as the previous time step, individual sensors within each sensor group are randomly matched to increase the diversity of the sensor groups, thereby improving the system's anti-interference capability. Different grouping combinations help discover more potential abnormal patterns and improve the comprehensiveness of anomaly detection. This prevents the system from over-relying on a specific grouping method, thus avoiding overfitting problems. This allows the system to maintain good performance even when facing new or unseen data. Maintaining the same number of groups as the previous time step ensures that the system maintains consistent monitoring density and coverage across different time periods. This helps maintain the overall balance of the system and avoids monitoring blind spots or redundancy caused by changes in the number of groups. The same number of groups ensures the comparability of data across different time periods, facilitating cross-time period data comparison and analysis. This is very helpful for long-term trend analysis and historical data backtracking.

[0028] Then, for each abnormal sensor, the data correlation degree between it and all normal sensors in each sensor group is retrieved. For example, if the abnormal sensor is A and the normal sensors in the sensor group are a, b, and c, then when retrieving the corresponding data correlation degree, the correlation degree between the abnormal sensor A and the normal sensors a, b, and c will be retrieved. By quantifying the relationship between normal and abnormal sensors, a foundation is provided for subsequent analysis.

[0029] Then, by combining the basic data of the underground roadway (such as geological structure and ventilation conditions) and the specific installation location of the sensors, the correlation of the retrieved data is dynamically adjusted, which greatly improves the authenticity and reliability of the adjusted data correlation and makes it more in line with the actual situation.

[0030] Furthermore, the prediction module is also used to retrieve the second monitoring and early warning prediction model corresponding to the normal sensor, and input the sensor data corresponding to the normal sensor as input data into the second monitoring and early warning prediction model, and output the second prediction result corresponding to the underground roadway at the current time.

[0031] The server also includes an analysis module, which is used to calculate the latest corrected prediction result based on the first and second prediction results corresponding to the current underground roadway and the weight values ​​corresponding to each prediction structure, according to the output.

[0032] The early warning module is also used to determine whether there is danger in the underground roadway at the current moment based on the latest corrected prediction results. If so, it sends a second early warning message to the management terminal.

[0033] Beneficial effects: In this scheme, by combining the first and second prediction results and using normal data to correct abnormal data, the probability of false alarms and missed alarms is effectively reduced, and the accuracy of prediction is significantly improved. This multi-model fusion method can more comprehensively reflect the actual environmental conditions of underground roadways.

[0034] The second prediction result, serving as an independent verification method, can effectively detect anomalies in the first prediction result, enhancing the robustness and reliability of the system. This dual verification mechanism ensures the stability and credibility of the prediction results. Furthermore, through correction calculations, the system can effectively reduce false alarms and false negatives caused by errors in a single model, improving the overall reliability of the system.

[0035] Furthermore, the basic data of the tunnel includes tunnel geological data and tunnel data, and the tunnel data includes tunnel structure data, tunnel support data and tunnel size data.

[0036] Furthermore, the server also includes:

[0037] The report generation module is used to generate a corresponding early warning and maintenance report and feed it back to the management terminal when the judgment result indicates that there is a danger in the underground roadway at the current time. The early warning and maintenance report includes the current early warning level and the corresponding maintenance solution.

[0038] Beneficial effects: When the system determines that there is a danger in the underground roadway at the current moment, the report generation module can instantly generate an early warning maintenance report and quickly feed it back to the management end. This instant response mechanism greatly shortens the time from risk detection to taking measures, improves the efficiency of emergency response, and the early warning maintenance report not only includes the current warning level, but also provides detailed maintenance solutions, enabling managers to quickly understand the specific situation and take appropriate measures.

[0039] This invention also provides a monitoring and early warning processing method for downhole construction sites, which, using the aforementioned monitoring and early warning processing system for downhole construction sites, includes the following steps:

[0040] S1. At the current moment, collect sensor data from each sensor located at each sensor point in a certain monitoring group within the underground roadway, based on a pre-defined sensor point setting scheme.

[0041] S2. Based on the collected sensor data and the number of sensor points in the corresponding sensor point setting scheme, determine in turn whether there is any abnormality in the sensor data of the corresponding sensor at each sensor point at the current time. If so, determine that the corresponding sensor is an abnormal sensor; if not, determine that the corresponding sensor is a normal sensor. Continue until the judgment of all sensors is completed, forming the corresponding abnormal sensor set and normal sensor set.

[0042] S3. Based on the sensor data of the normal sensors corresponding to the normal sensor set at the current time, and based on the preset data correlation selection strategy, select the data correlation set corresponding to each abnormal sensor in the abnormal sensor set, and calculate the sensor data corresponding to the abnormal sensor based on each data correlation set corresponding to each abnormal sensor and the sensor data of the normal sensor corresponding to the data correlation.

[0043] S4. After completing the calculation of sensor data of all abnormal sensors, retrieve the first monitoring and early warning model corresponding to the abnormal sensor and the normal sensor, and input the sensor data corresponding to the abnormal sensor and the sensor data corresponding to the normal sensor as input data into the corresponding first monitoring and early warning prediction model, and output the first prediction result corresponding to the underground roadway at the current moment.

[0044] S5. Based on the first prediction result corresponding to the underground roadway at the current moment, determine whether there is danger in the underground roadway at the current moment. If so, send the first warning information to the management terminal. Attached Figure Description

[0045] Figure 1This is a logic block diagram of the monitoring and early warning processing system for underground construction sites in Embodiment 1 of the present invention;

[0046] Figure 2 This is a flowchart of a monitoring and early warning processing method for underground construction sites according to Embodiment 1 of the present invention. Detailed Implementation

[0047] The following detailed description illustrates the specific implementation method:

[0048] Example 1

[0049] Monitoring, early warning, and processing systems and methods for downhole construction sites are basically as follows: Figure 1 As shown, it includes the server and management sides;

[0050] The server includes:

[0051] The data acquisition module is used to collect sensor data from various sensors located at different sensor positions within a pre-defined sensor placement scheme in the underground roadway at the current moment. In this embodiment, the sensor types include temperature sensors, humidity sensors, gas concentration sensors, anchor stress gauges, surface strain gauges, axial force gauges, vibrating wire data acquisition instruments, crack gauges, rebar gauges, earth pressure cells, and multi-point displacement gauges. Different sensors correspond to different installation locations and methods. For example, when installing a multi-point displacement gauge, the calibrated gauge is placed into a pre-drilled hole (with a φ90-110 drill bit diameter) at a depth of 4 meters. After drilling, the hole is thoroughly cleaned with pressurized water before instrument installation. The grouting casing is anchored around the opening with cement mortar or epoxy anchoring agent, with the outer side of the casing flush with the hole opening. During installation, the sensor must be secured to prevent it from falling and causing a safety accident. The installation location is set around the initial support rock. The installation method for the earth pressure gauge is as follows: The earth pressure gauge is placed tightly against the surface of the primary support rock and the secondary support structure, and fixed using a special fixing device. During installation, care must be taken to ensure that no gaps form between the pressure surface of the earth pressure gauge and the soil, and water accumulation caused by soil separation between the instrument and the soil should be minimized. Installation locations: primary support rock, surface of the primary support structure, surface of the secondary support structure.

[0052] In this embodiment, the sensors in the underground roadway are arranged according to a pre-defined sensor location plan. For the sensors deployed at each sensor location of a specific monitoring group in the underground roadway, their sensor data are accurately collected, thereby providing basic data support and guarantee for subsequent monitoring and early warning work.

[0053] The identification module is used to determine whether there is any abnormality in the sensor data of the corresponding sensor at each sensor location at the current time, based on the collected sensor data and the number of sensor points in the corresponding sensor location setting scheme. If so, the corresponding sensor is identified as an abnormal sensor; otherwise, the corresponding sensor is identified as a normal sensor. This process continues until all sensors are identified, forming their respective sets of abnormal and normal sensors.

[0054] The calculation module is used to select the data correlation set corresponding to each abnormal sensor in the abnormal sensor set based on the sensor data of the normal sensor set corresponding to the normal sensor set at the current time and a preset data correlation selection strategy. Based on the data correlation set corresponding to each abnormal sensor and the sensor data of the normal sensor corresponding to the data correlation, the module calculates the sensor data corresponding to the abnormal sensor.

[0055] The preset data correlation selection strategy includes the following steps:

[0056] S100. Based on the normal sensor set and the abnormal sensor set at the current moment, calculate the ratio of the number of sensors between the normal sensor set and the abnormal sensor set;

[0057] S200. If the sensor quantity ratio is greater than or equal to the preset quantity ratio, then based on the grouping situation corresponding to the previous moment, the normal sensors in the normal sensor set are grouped to form various sensor groups; the grouping situation includes the number of groups and the sensors corresponding to each group.

[0058] S300. If the number ratio of sensors is less than the preset number ratio, then according to each normal sensor in the normal sensor set and the number of groups corresponding to the previous moment, each normal sensor in the normal sensor set is randomly selected and combined to form a sensor group with the same number of groups as the previous moment.

[0059] S400. Based on each sensor in each sensor group, retrieve the data correlation degree between each abnormal sensor in the abnormal sensor set and all normal sensors in each sensor group.

[0060] S500: Obtain the basic roadway data of the corresponding monitoring group in the underground roadway and the installation location data of the monitoring group in the underground roadway, and dynamically adjust the correlation of the retrieved data based on the basic roadway data and the installation location data; In this embodiment, the basic roadway data includes roadway geological data and roadway data, and the roadway data includes roadway structure data, roadway support data and roadway size data.

[0061] S500. Based on the data correlation degree between each abnormal sensor in the dynamically adjusted abnormal sensor set and all normal sensors in each sensor group, as well as the sensor data corresponding to all normal sensors in the sensor group, calculate the sensor data corresponding to each abnormal sensor in the abnormal sensor set.

[0062] For example, a monitoring group includes 13 different sensors. At the current moment, there are 10 normal sensors and 2 abnormal sensors in the underground roadway. Each sensor has corresponding sensor data. Therefore, the ratio of the number of sensors to the number of abnormal sensors is... The corresponding preset quantity ratio is 4.

[0063] Since the sensor data ratio B = 5 is less than the preset quantity ratio of 4, it is necessary to group the data based on the grouping situation at the previous time step. Assume there were 3 groups at the previous time step: one group included sensors a, b, and c; another group included sensors d, e, and f; and the last group included sensors h, i, j, and k. The current grouping will follow the same pattern.

[0064] Assuming there are 6 normal sensors and 6 abnormal sensors, what is the ratio of the number of sensors? Since the number of sensors is less than the preset number ratio of 4, random grouping is required. The number of groups is the same as the number of groups in the previous moment. Assuming that the number of groups in the previous moment was also 3, then the 6 normal sensors can be randomly selected and placed into three groups. For example, the final sensor group 1 includes sensor 1 and sensor 2, sensor group 2 includes sensor 3, and sensor group 3 includes sensor 4, sensor 5 and sensor 6.

[0065] After grouping is completed, the data correlation degree will be retrieved. During retrieval, the abnormal sensor A will retrieve the data correlation degree corresponding to the abnormal sensor A with sensor group 1, sensor group 2 and sensor group 3. The data correlation degree corresponding to each group is the common data correlation degree between all sensors in the group and the abnormal sensor A.

[0066] The prediction module is used to retrieve the first monitoring and early warning models corresponding to the abnormal sensors and normal sensors after completing the calculation of the sensor data of all abnormal sensors, and input the sensor data corresponding to the abnormal sensors and the sensor data corresponding to the normal sensors as input data into the corresponding first monitoring and early warning prediction model, and output the first prediction result corresponding to the underground roadway at the current moment.

[0067] The prediction module is further configured to retrieve the second monitoring and early warning prediction model corresponding to the normal sensors, and input the sensor data corresponding to the normal sensors into the second monitoring and early warning prediction model, outputting the second prediction result corresponding to the underground roadway at the current moment. In this embodiment, the first prediction result is formed by inputting the sensor data of the abnormal sensors and the sensor data of the normal sensors into the corresponding model after supplementing the sensor data of the abnormal sensors. The second prediction result is formed by inputting the sensor data of the normal sensors into the corresponding model. Through the analysis and data fusion of the two prediction results, a new prediction result is formed, greatly improving the authenticity and reliability of the prediction result. In this embodiment, both the first and second monitoring and early warning prediction models are constructed using BP neural network technology. Furthermore, multiple second monitoring and early warning prediction models are set, and different second monitoring and early warning prediction models correspond to different sensors.

[0068] For example, if there are 5 sensors, the model will be constructed based on the sensor data corresponding to these 5 sensors, using multiple sensor combinations to form a second monitoring and early warning prediction model. For instance, the sensor group formed by sensors aa, ba, ca, and da corresponds to the second monitoring and early warning model AA. Specifically, a three-layer BP neural network model is first constructed, including an input layer, hidden layers, and an output layer. In this embodiment, the sensor data corresponding to sensors aa, ba, ca, and da are used as the input to the input layer, therefore the input layer has 4 nodes. The output is the corresponding second prediction result, therefore it has 1 node. For the hidden layer, this embodiment uses the following formula to determine the number of hidden layer nodes: Where Y is the number of nodes in the hidden layer, n is the number of nodes in the input layer, m is the number of nodes in the output layer, and a is a number between 1 and 10, which is taken as 4 in this embodiment. Therefore, there are a total of 6 hidden layer nodes. Backpropagation (BP) neural networks typically use the sigmoid differentiable function and linear functions as the network's activation functions. This paper selects the sigmoid tangent function (tansig) as the activation function for the hidden layer neurons. The prediction model selects the sigmoid logarithmic function (tansig) as the activation function for the output layer neurons.

[0069] The server also includes an analysis module, which is used to calculate the latest corrected prediction result based on the first and second prediction results corresponding to the current underground roadway and the weight values ​​corresponding to each prediction structure, according to the output.

[0070] The early warning module is also used to determine whether there is danger in the underground roadway at the current moment based on the latest corrected prediction results. If so, it sends a second early warning message to the management terminal.

[0071] The early warning module is used to determine whether there is a hazard in the underground roadway at the current moment based on the first prediction result corresponding to the current moment. If so, it sends a first early warning message to the management terminal. In this embodiment, based on the first prediction result corresponding to the underground roadway at the current moment, it makes a more in-depth judgment on whether the underground roadway is in a dangerous state at that moment. If the judgment result is that there is a hazard, the first early warning message is immediately and accurately pushed to the management terminal to achieve timely notification and efficient control of risks.

[0072] The server also includes:

[0073] The report generation module is used to generate a corresponding early warning and maintenance report and feed it back to the management terminal when the judgment result indicates that there is a danger in the underground roadway at the current time. The early warning and maintenance report includes the current early warning level and the corresponding maintenance solution.

[0074] like Figure 2 As shown, this embodiment also discloses a monitoring and early warning processing method for underground construction sites. Using the above-mentioned monitoring and early warning processing system for underground construction sites, the method includes the following steps:

[0075] S1. At the current moment, collect sensor data from each sensor located at each sensor point in a certain monitoring group within the underground roadway, based on a pre-defined sensor point setting scheme.

[0076] S2. Based on the collected sensor data and the number of sensor points in the corresponding sensor point setting scheme, determine in turn whether there is any abnormality in the sensor data of the corresponding sensor at each sensor point at the current time. If so, determine that the corresponding sensor is an abnormal sensor; if not, determine that the corresponding sensor is a normal sensor. Continue until the judgment of all sensors is completed, forming the corresponding abnormal sensor set and normal sensor set.

[0077] S3. Based on the sensor data of the normal sensors corresponding to the normal sensor set at the current time, and based on the preset data correlation selection strategy, select the data correlation set corresponding to each abnormal sensor in the abnormal sensor set, and calculate the sensor data corresponding to the abnormal sensor based on each data correlation set corresponding to each abnormal sensor and the sensor data of the normal sensor corresponding to the data correlation.

[0078] S4. After completing the calculation of sensor data of all abnormal sensors, retrieve the first monitoring and early warning model corresponding to the abnormal sensor and the normal sensor, and input the sensor data corresponding to the abnormal sensor and the sensor data corresponding to the normal sensor as input data into the corresponding first monitoring and early warning prediction model, and output the first prediction result corresponding to the underground roadway at the current moment.

[0079] S5. Based on the first prediction result corresponding to the underground roadway at the current moment, determine whether there is danger in the underground roadway at the current moment. If so, send the first warning information to the management terminal.

[0080] The above descriptions are merely embodiments of the present invention. Commonly known structures and characteristics are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are aware of all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, based on the guidance provided in this application, improve and implement this solution in combination with their own capabilities. Some typical well-known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of the present invention. These should also be considered within the scope of protection of the present invention, and will not affect the effectiveness of the implementation of the present invention or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims, and the specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A monitoring and early warning system for use in underground construction sites, characterized in that: Including server-side and management-side; The server includes: The data acquisition module is used to collect sensor data from each sensor at each sensor point in a certain monitoring group within the underground roadway at the current moment, based on a pre-defined sensor point setting scheme. The identification module is used to determine whether there is any abnormality in the sensor data of the corresponding sensor at each sensor location at the current time, based on the collected sensor data and the number of sensor points in the corresponding sensor location setting scheme. If so, the corresponding sensor is identified as an abnormal sensor; otherwise, the corresponding sensor is identified as a normal sensor. This process continues until all sensors are identified, forming their respective sets of abnormal and normal sensors. The calculation module is used to select the data correlation set corresponding to each abnormal sensor in the abnormal sensor set based on the sensor data of the normal sensor set corresponding to the normal sensor set at the current time and a preset data correlation selection strategy. Based on the data correlation set corresponding to each abnormal sensor and the sensor data of the normal sensor corresponding to the data correlation, the module calculates the sensor data corresponding to the abnormal sensor. The prediction module is used to retrieve the first monitoring and early warning models corresponding to the abnormal sensors and normal sensors after completing the calculation of the sensor data of all abnormal sensors, and input the sensor data corresponding to the abnormal sensors and the sensor data corresponding to the normal sensors as input data into the corresponding first monitoring and early warning prediction model, and output the first prediction result corresponding to the underground roadway at the current moment. The early warning module is used to determine whether there is any danger in the underground roadway at the current moment based on the first prediction result corresponding to the current moment. If so, it sends the first early warning information to the management terminal.

2. The monitoring and early warning system for downhole construction sites according to claim 1, characterized in that: The preset data correlation selection strategy includes the following steps: S100. Based on the normal sensor set and the abnormal sensor set at the current moment, calculate the ratio of the number of sensors between the normal sensor set and the abnormal sensor set; S200. If the sensor quantity ratio is greater than or equal to the preset quantity ratio, then based on the grouping situation corresponding to the previous moment, the normal sensors in the normal sensor set are grouped to form various sensor groups; the grouping situation includes the number of groups and the sensors corresponding to each group. S300. If the number ratio of sensors is less than the preset number ratio, then according to each normal sensor in the normal sensor set and the number of groups corresponding to the previous moment, each normal sensor in the normal sensor set is randomly selected and combined to form a sensor group with the same number of groups as the previous moment. S400. Based on each sensor in each sensor group, retrieve the data correlation degree between each abnormal sensor in the abnormal sensor set and all normal sensors in each sensor group. S500: Obtain the basic roadway data of the corresponding monitoring group in the underground roadway and the installation location data of the monitoring group in the underground roadway, and dynamically adjust the correlation of the retrieved data based on the basic roadway data and the installation location data. S500. Based on the data correlation degree between each abnormal sensor in the dynamically adjusted abnormal sensor set and all normal sensors in each sensor group, as well as the sensor data corresponding to all normal sensors in the sensor group, calculate the sensor data corresponding to each abnormal sensor in the abnormal sensor set.

3. The monitoring and early warning system for downhole construction sites according to claim 2, characterized in that: The prediction module is also used to retrieve the second monitoring and early warning prediction model corresponding to the normal sensor, and input the sensor data corresponding to the normal sensor as input data into the second monitoring and early warning prediction model, and output the second prediction result corresponding to the underground roadway at the current time. The server also includes an analysis module, which is used to calculate the latest corrected prediction result based on the first and second prediction results corresponding to the current underground roadway and the weight values ​​corresponding to each prediction structure, according to the output. The early warning module is also used to determine whether there is danger in the underground roadway at the current moment based on the latest corrected prediction results. If so, it sends a second early warning message to the management terminal.

4. The monitoring and early warning processing system for downhole construction sites according to claim 3, characterized in that: The basic data of the tunnel includes tunnel geological data and tunnel data, which includes tunnel structure data, tunnel support data, and tunnel size data.

5. The monitoring and early warning system for downhole construction sites according to claim 4, characterized in that: The server also includes: The report generation module is used to generate a corresponding early warning and maintenance report and feed it back to the management terminal when the judgment result indicates that there is a danger in the underground roadway at the current time. The early warning and maintenance report includes the current early warning level and the corresponding maintenance solution.

6. A monitoring and early warning processing method for underground construction sites, using the monitoring and early warning processing system for underground construction sites according to any one of claims 1 to 5, characterized in that: Includes the following steps: S1. At the current moment, collect sensor data from each sensor located at each sensor point in a certain monitoring group within the underground roadway, based on a pre-defined sensor point setting scheme. S2. Based on the collected sensor data and the number of sensor points in the corresponding sensor point setting scheme, determine in turn whether there is any abnormality in the sensor data of the corresponding sensor at each sensor point at the current time. If yes, the corresponding sensor is determined to be an abnormal sensor; otherwise, the corresponding sensor is a normal sensor. This continues until all sensors have been evaluated, resulting in corresponding sets of abnormal and normal sensors. S3. Based on the sensor data of the normal sensors corresponding to the normal sensor set at the current time, and based on the preset data correlation selection strategy, select the data correlation set corresponding to each abnormal sensor in the abnormal sensor set, and calculate the sensor data corresponding to the abnormal sensor based on each data correlation set corresponding to each abnormal sensor and the sensor data of the normal sensor corresponding to the data correlation. S4. After completing the calculation of sensor data of all abnormal sensors, retrieve the first monitoring and early warning model corresponding to the abnormal sensor and the normal sensor, and input the sensor data corresponding to the abnormal sensor and the sensor data corresponding to the normal sensor as input data into the corresponding first monitoring and early warning prediction model, and output the first prediction result corresponding to the underground roadway at the current moment. S5. Based on the first prediction result corresponding to the underground roadway at the current moment, determine whether there is danger in the underground roadway at the current moment. If so, send the first warning information to the management terminal.

Images