Method and system for personnel regulation in limited space of underground engineering

Abstract

This application relates to the field of personnel management technology in underground spaces, specifically to a method and system for controlling personnel in confined spaces in underground engineering projects. By deploying multiple infrared detection modules at the entrance, non-contact data on personnel entry and exit is collected. Combined with a pre-set recognition model, entry behavior is accurately identified and the number of personnel is updated in real time. Based on this data, dynamic spatial information is generated. During the blasting preparation phase, the system automatically verifies the clearing status and generates blasting instructions. During the operation phase, it monitors whether limits are exceeded in real time and triggers evacuation commands. The system also dynamically updates the clearing and over-limit judgment criteria according to the operation conditions and introduces a sliding window to calculate the average personnel flow, thereby correcting the over-limit threshold and achieving adaptive control. This application effectively solves the problems of inaccurate traditional manual statistics, delayed early warning, and rigid thresholds, improving the safety, intelligence, and robustness of personnel management in underground engineering projects.

Description

Technical Field

[0001] This application relates to the field of personnel management technology in underground spaces, specifically to methods and systems for personnel control in confined spaces of underground engineering projects. Background Technology

[0002] Underground engineering projects (such as tunnels, mines, underground utility tunnels, and subway construction sections) are typically confined space environments, characterized by limited access, poor ventilation, insufficient lighting, and communication difficulties. In the event of a safety accident (such as accidental injury during blasting, overcrowding or suffocation due to exceeding personnel limits, or delayed emergency evacuation), significant casualties are highly likely. Therefore, real-time, accurate, and dynamic monitoring and control of the number of personnel inside underground engineering projects is a core element in ensuring operational safety.

[0003] Traditional underground engineering personnel management relies heavily on manual sign-in, clock-in, or manual interpretation of video surveillance. However, manual statistics suffer from low efficiency, easy omissions, and inability to update in real time; while conventional video surveillance is easily affected by factors such as dust, humidity, and insufficient light in the complex underground environment, making it difficult to guarantee accuracy. In addition, existing systems generally use fixed thresholds to determine whether personnel have exceeded limits or whether site clearance has been completed, lacking the ability to adapt to different work stages (such as blasting preparation, normal tunneling, support work, etc.), different project types, and fluctuations in personnel flow. This leads to frequent false alarms during peak periods or overly conservative approaches during low-risk periods, affecting construction efficiency and the accuracy of safety responses.

[0004] Especially before blasting operations, it is essential to ensure that no personnel remain in the underground space. Traditional methods rely on manual counting and verbal confirmation, which carries the risk of human negligence or delays in information transmission, making it difficult to meet the "zero tolerance" safety requirements for high-risk operations. Furthermore, in routine underground operations, if the personnel density exceeds the space's carrying capacity or emergency evacuation capabilities, it could lead to mass casualties. Therefore, a technological means is urgently needed that can sense situations in real time, make intelligent judgments, and automatically trigger early warnings or control commands.

[0005] In recent years, the development of infrared sensing, edge computing, and lightweight artificial intelligence models has provided new ideas for non-contact personnel detection. However, how to effectively integrate multi-source infrared data, construct recognition models suitable for complex underground scenarios, and on this basis realize dynamic control mechanisms linked to operational conditions (such as site clearance verification, over-limit warning, and threshold adaptive correction) remains a challenge that current technology has not yet fully solved.

[0006] Therefore, there is an urgent need for an intelligent control method that can automatically count, dynamically monitor, and manage the safety of personnel in confined spaces in underground engineering, and has the ability to adapt to different working conditions. This method can overcome the shortcomings of existing technologies, such as inaccurate statistics, delayed response, rigid thresholds, and poor environmental adaptability, and comprehensively improve the safety, reliability, and intelligence level of underground engineering operations. Summary of the Invention

[0007] In view of this, this application provides a method and system for personnel control in confined spaces of underground engineering projects, which can realize automatic statistics, dynamic monitoring, and closed-loop safety management of personnel in confined spaces of underground engineering projects, and has the ability to adapt to working conditions. This overcomes the defects of existing technologies such as inaccurate statistics, delayed response, rigid thresholds, and poor environmental adaptability, and comprehensively improves the safety, reliability and intelligence level of underground engineering operations.

[0008] In a first aspect, this application provides a method for personnel control in a confined space of an underground engineering project, comprising: obtaining entrance monitoring data at the entrance of the underground engineering project based on multiple infrared detection modules; calling a preset recognition model to identify personnel entry and exit data based on the entrance monitoring data, and updating real-time personnel data based on the personnel entry and exit data; obtaining spatial dynamic data of the internal space of the underground engineering project based on the real-time personnel data; in the blasting preparation stage, if the spatial dynamic data meets the personnel clearing conditions and continues for a first preset duration, generating a blasting instruction; in the underground operation stage, if the spatial dynamic data reaches the personnel over-limit condition, generating an evacuation instruction; updating the corresponding personnel clearing conditions and personnel over-limit conditions periodically according to different application conditions; setting a sliding window based on the current time, calculating the average personnel flow in the sliding window, and obtaining a corresponding first correction parameter based on the average personnel flow; the average personnel flow and the first correction condition are proportional; and correcting the personnel over-limit condition based on the first correction parameter.

[0009] In conjunction with the first aspect, in one possible implementation, obtaining the entrance monitoring data at the entrance of the underground project based on multiple infrared detection modules includes: receiving individual infrared data detected by each of the infrared detection modules at the entrance of the underground project; and summing up the multiple individual infrared data to obtain the entrance monitoring data.

[0010] In conjunction with the first aspect, in one possible implementation, the step of calling a preset recognition model to identify personnel entry and exit data based on the entrance monitoring data, and updating real-time personnel data based on the personnel entry and exit data, includes: extracting the 35℃-38℃ thermal radiation area of ​​personnel from the single infrared data; marking target feature points in the personnel thermal radiation area; locating the relative spatial coordinates between the target feature point and the infrared detection module based on the module spatial coordinates of the infrared detection module; generating a relative line connecting the module spatial coordinates and the relative spatial coordinates and obtaining the corresponding relative spatial angle; calculating the absolute spatial coordinates of the target feature point based on multiple sets of relative spatial angles and multiple module spatial coordinates; tracking the movement trajectory of the absolute spatial coordinates to obtain the personnel entry and exit data; and based on the personnel entry and exit data, incrementing the number of people in the venue by one if someone enters and decrementing the number of people in the venue by one if someone exits.

[0011] In conjunction with the first aspect, in one possible implementation, after calculating the absolute spatial coordinates of the target feature point based on multiple sets of relative spatial angles and multiple module spatial coordinates, the method further includes: performing data filtering on the absolute spatial coordinates; obtaining fusion vector parameters based on the absolute spatial coordinates and their movement speed; obtaining the current predicted coordinates based on the fusion vector parameters; acquiring the current observed coordinates and correcting the current predicted coordinates, and outputting the corrected absolute spatial coordinates.

[0012] In conjunction with the first aspect, in one possible implementation, the step of tracking the motion trajectory of the absolute spatial coordinates to obtain the personnel entry and exit data includes: refreshing the absolute spatial coordinates based on a preset frequency; and connecting consecutive absolute spatial coordinates in series according to timestamps to generate the motion trajectory.

[0013] In conjunction with the first aspect, in one possible implementation, the step of setting a sliding window based on the current time, calculating the average pedestrian flow within the sliding window, and obtaining a corresponding first correction parameter based on the average pedestrian flow includes: setting the sliding window back a preset time period with the current time as a reference point; summarizing the total pedestrian flow within the sliding window; calculating the average pedestrian flow corresponding to a set unit time period based on the total pedestrian flow and the preset time period; setting a corresponding benchmark reference flow based on the spatial area of ​​the underground space; if the average pedestrian flow is less than or equal to the benchmark reference flow, the first correction parameter is 1; if the average pedestrian flow is greater than the benchmark reference flow but less than twice the benchmark reference flow, the value range of the first correction parameter is greater than 1 and less than 2; if the average pedestrian flow is greater than or equal to twice the benchmark reference flow, the first correction parameter is 2.

[0014] In conjunction with the first aspect, one possible implementation also includes: setting a sensitivity coefficient; and adjusting the preset duration based on the sensitivity coefficient.

[0015] In conjunction with the first aspect, one possible implementation further includes: obtaining the safety requirement level of the underground space; obtaining a corresponding second correction coefficient based on the safety requirement level; the higher the safety requirement level, the smaller the second correction coefficient; and performing a secondary correction on the personnel over-limit working conditions based on the second correction coefficient.

[0016] In conjunction with the first aspect, in one possible implementation, the safety requirement levels include a general level, a dangerous level, and an extremely dangerous level, with the second correction coefficients corresponding to the general level, the dangerous level, and the extremely dangerous level being 1, 0.6, and 0.3, respectively.

[0017] Secondly, this application provides a personnel control system for confined spaces in underground engineering projects, comprising: a monitoring module configured to: obtain entrance monitoring data at the entrance of the underground engineering project based on multiple infrared detection modules; a space dynamic refresh module, communicatively connected to the monitoring module, the space dynamic refresh module being configured to: call a preset recognition model to identify personnel entry and exit data based on the entrance monitoring data, and update real-time personnel data based on the personnel entry and exit data; and obtain spatial dynamic data of the internal space of the underground engineering project based on the real-time personnel data; and an instruction generation module, communicatively connected to the space dynamic refresh module, the instruction generation module being configured to: during the blasting preparation stage, if the... If the spatial dynamic data meets the personnel clearing conditions and continues for a first preset duration, a blasting instruction is generated; during underground operations, if the spatial dynamic data reaches the personnel over-limit condition, an evacuation instruction is generated; a correction module, communicatively connected to the instruction generation module, is configured to: periodically update the corresponding personnel clearing conditions and personnel over-limit conditions according to different application conditions; set a sliding window based on the current time, calculate the average flow of people in the sliding window, and obtain a corresponding first correction parameter based on the average flow of people; the average flow of people is proportional to the first correction condition; and correct the personnel over-limit condition based on the first correction parameter.

[0018] The personnel control method for confined spaces in underground engineering provided in this application achieves closed-loop management of the entire process, including personnel statistics, dynamic monitoring, safety control, and adaptive optimization of working conditions. Through multiple infrared detection modules, it achieves non-contact, all-weather stable monitoring at entrances, avoiding interference from the complex underground environment and providing reliable data for subsequent statistics. It accurately identifies personnel entry and exit information using a preset recognition model, updates personnel data in real time, and transforms it into dynamic spatial data, breaking through the limitations of traditional manual statistics. For blasting and underground operation scenarios, it automatically generates blasting instructions after site clearance verification and evacuation instructions when exceeding limits, reducing operational safety risks. Furthermore, this application improves system adaptability and robustness by periodically updating working condition judgment standards and dynamically correcting over-limit working conditions through a sliding window, effectively solving pain points such as inaccurate personnel statistics and delayed early warning in underground confined spaces, and significantly improving operational safety, efficiency, and intelligence. Attached Figure Description

[0019] Figure 1 The diagram shows the steps of a method for controlling personnel in a confined space in an underground engineering project, according to an embodiment of this application.

[0020] Figure 2 The diagram shows the steps involved in obtaining entrance monitoring data based on infrared data.

[0021] Figure 3 The diagram shows the steps involved in calculating the number of attendees based on spatial coordinates.

[0022] Figure 4 The diagram shows the steps involved in correcting spatial coordinates.

[0023] Figure 5 The diagram shows the steps involved in generating a motion trajectory.

[0024] Figure 6 The diagram shows the steps of calculating the first correction parameter based on a sliding window.

[0025] Figure 7 The diagram shows the steps for adjusting the preset duration of a sliding window.

[0026] Figure 8 The diagram shows the steps of the method for correcting over-limit conditions in a second correction.

[0027] Figure 9 The diagram shown is a schematic diagram of the system structure of a personnel control system for a confined space in an underground engineering project, according to an embodiment of this application. Detailed Implementation

[0028] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0029] An exemplary method for personnel control in confined spaces of underground engineering projects is as follows: Figure 1 The diagram shown is a schematic representation of the steps in a method for controlling personnel in a confined space in an underground engineering project, according to an embodiment of this application. This application provides a method for controlling personnel in a confined space in an underground engineering project, such as... Figure 1 As shown, the method includes: Step 110: Obtain entrance monitoring data at the entrance of the underground project based on multiple infrared detection modules.

[0030] In this step, multiple infrared detection modules are used to monitor the entrance to the underground project. These modules identify whether someone is passing through by detecting the heat emitted by the human body. This enables non-contact, all-weather, and stable data collection on personnel passage at the underground project entrance, providing reliable raw input for subsequent identification.

[0031] Step 120: Call the preset recognition model to identify personnel entry and exit data based on the entrance monitoring data, and update the real-time personnel data based on the personnel entry and exit data.

[0032] In this step, a pre-defined recognition model is applied to process the entrance monitoring data to identify whether personnel are entering or leaving the infrared data, distinguishing between entry and exit, and updating the real-time personnel database. This enables dynamic monitoring of the number of people inside the underground project, ensuring the timeliness and accuracy of the data.

[0033] Step 130: Obtain the spatial dynamic data of the internal space of the underground project based on real-time personnel data.

[0034] In this step, the spatial dynamics data inside the underground project are calculated based on real-time updated personnel data, including but not limited to the number of people and personnel density.

[0035] Step 140: During the blasting preparation stage, if the spatial dynamic data matches the personnel clearing conditions and continues for a first preset duration, a blasting instruction is generated.

[0036] In this step, the spatial dynamic data is used to determine thresholds during the blasting preparation stage: if the personnel clearing conditions are met (e.g., number of people = 0) and this condition is maintained for a first preset duration, the clearing is deemed reliable, and the automatically generated blasting instruction is used to indicate that the underground project can be blasted. This step achieves automatic clearing verification before blasting, avoiding safety risks caused by personnel remaining on site, and ensuring the compliance and safety of blasting operations.

[0037] Step 150: During the underground operation phase, if the spatial dynamic data reaches the condition of exceeding the personnel limit, an evacuation instruction will be generated.

[0038] In this step, during the underground operation phase, dynamic spatial data is continuously compared with personnel overload conditions. Personnel overload conditions are defined as a set number of people in the underground space; once the threshold is reached, overload is determined, and an evacuation command is automatically issued. This achieves real-time early warning and forced evacuation triggering for confined spaces, preventing risks such as overcrowding, suffocation, and difficulty in emergency evacuation.

[0039] Step 160: Update the corresponding personnel clearing conditions and personnel over-limit conditions periodically according to different application conditions.

[0040] This step involves various application scenarios, including multiple work stages, procedures, and project types. It periodically updates the personnel clearing status for site clearance and the personnel exceeding limits status for limit determination. This step adapts the judgment criteria to different work scenarios, improving the applicability and accuracy of statistics and control.

[0041] Step 170: Based on the current time, set a sliding window, calculate the average pedestrian flow in the sliding window, and obtain the corresponding first correction parameter based on the average pedestrian flow; the average pedestrian flow and the first correction parameter are directly proportional.

[0042] In this step, a sliding time window is set based on the current moment, and the average pedestrian flow within the window is statistically analyzed to establish a positive correlation between the average pedestrian flow and the first correction parameter. This step uses historical flow characteristics to quantify the intensity of pedestrian flow fluctuations, providing a quantitative basis for dynamic correction of exceeding the threshold and avoiding the incompatibility of fixed thresholds with peak flow.

[0043] Step 180: Correct the personnel over-limit working conditions based on the first correction parameter.

[0044] In this step, the personnel over-limit threshold, i.e., the personnel over-limit working condition, is dynamically adjusted using the first correction parameter. The larger the passenger flow, the more closely the corrected threshold matches the actual carrying and evacuation capacity. This step achieves adaptive correction of the personnel over-limit threshold, preventing overload and avoiding frequent false alarms during peak periods, thus improving the system's robustness and practicality.

[0045] The personnel control method for confined spaces in underground engineering provided in this embodiment achieves closed-loop management of the entire process, including personnel statistics, dynamic monitoring, safety control, and adaptive optimization of working conditions. Through multiple infrared detection modules, it achieves non-contact, all-weather stable monitoring of entrances, avoiding interference from the complex underground environment and providing reliable data for subsequent statistics. It accurately identifies personnel entry and exit information using a preset recognition model, updates personnel data in real time, and transforms it into dynamic spatial data, breaking through the limitations of traditional manual statistics. For blasting and underground operation scenarios, it automatically generates blasting instructions after site clearance verification and evacuation instructions when exceeding limits, reducing operational safety risks. Furthermore, this embodiment improves system adaptability and robustness by periodically updating working condition judgment standards and dynamically correcting exceeding-limit working conditions through a sliding window, effectively solving pain points such as inaccurate personnel statistics and delayed early warnings in underground confined spaces, significantly improving operational safety, efficiency, and intelligence.

[0046] Figure 2 The diagram illustrates the steps of a method for obtaining entrance monitoring data based on infrared data. In one embodiment, as shown... Figure 2 As shown, step 110 includes: Step 111: Receive individual infrared data detected by each infrared detection module at the entrance of the underground project.

[0047] Step 112: Summarize multiple individual infrared data to obtain entrance monitoring data.

[0048] In this embodiment, each infrared detection module independently collects individual infrared data generated by human thermal radiation within its field of view. Then, the system aggregates and merges the individual infrared data from multiple modules into unified entry monitoring data.

[0049] Figure 3 The diagram illustrates the steps of a method for calculating the number of attendees based on spatial coordinates. In one embodiment, as shown... Figure 3 As shown, step 120 includes: Step 121: Extract the thermal radiation area of ​​personnel at 35℃-38℃ from a single infrared data.

[0050] Step 122: Mark target feature points in the thermal radiation area of ​​personnel.

[0051] In this step, representative key points are selected within the thermal radiation area as target feature points, such as the center of the hot spot or the geometric center of the thermal radiation area, for subsequent spatial positioning and trajectory tracking.

[0052] Step 123: Based on the module spatial coordinates of the infrared detection module, locate the relative spatial coordinates between the corresponding target feature point and the infrared detection module.

[0053] In this step, based on the known installation location of each infrared detection module, i.e., the module's spatial coordinates, the infrared detection module may include an infrared ranging module to detect the distance between the target feature point and the infrared detection module, thereby obtaining their relative spatial coordinates.

[0054] Step 124: Generate the relative connection between the module space coordinates and the relative space coordinates, and obtain the corresponding relative space angle.

[0055] Step 125: Calculate the absolute spatial coordinates of the target feature point based on multiple sets of relative spatial angles and multiple module spatial coordinates.

[0056] In steps 124 and 125, the target feature point is denoted as P, and infrared detection modules A and B are set on both sides of the entrance. Using the principle of binocular vision matching, the relative spatial coordinates (u1, v1) of the target feature point P in the thermal image coordinate system of module A and the relative spatial coordinates (u2, v2) in the thermal image coordinate system of module B are obtained respectively. After connecting the lines, the relative spatial angles α and β are calculated. The relative spatial angle α is the angle between the line connecting points A and P and the line segment AB, and the relative spatial angle β is the angle between the line connecting points B and P and the line segment AB. Based on the triangle angle sum theorem and the sine theorem, given the distance L between points A and B, and the angles α and β, the absolute spatial coordinates (x, y) of point P relative to the origin can be calculated: x=[(L×cosα×cosβ)-(L×sinα×sinβ)] / 2; y=[(L×sinα×cosβ)+(L×cosα×sinβ)] / 2; Based on the above principle, coordinate calculation is automatically performed in real time, and the real-time absolute spatial coordinates of point P are output to complete a single positioning.

[0057] Step 126: Track the movement trajectory of the absolute spatial coordinates to obtain personnel entry and exit data.

[0058] In this step, the absolute spatial coordinates in continuous time frames are correlated temporally and fitted with trajectories to determine whether the target's movement direction is entering or leaving the underground engineering space, thereby generating personnel entry and exit event data.

[0059] Step 127: Based on the entry and exit data, if someone enters, the number of people in the venue increases by one; if someone exits, the number of people in the venue decreases by one.

[0060] This embodiment achieves high-precision and high-reliability identification and statistics of personnel entry and exit behavior at underground engineering entrances by employing multiple infrared detection modules working collaboratively, combined with thermal radiation feature extraction and spatial geometric positioning technologies. The target area is extracted based on the typical human body thermal radiation temperature range of 35℃–38℃, effectively filtering out environmental thermal interference and significantly improving the accuracy of personnel detection. Multiple infrared detection modules (such as dual-sided deployment) acquire multi-view observation data of target feature points, and the absolute spatial coordinates of the target are calculated in real time using the triangulation principle. By performing time-series tracking and motion direction analysis of the continuous trajectory of the target's absolute spatial coordinates, "entry" and "exit" behaviors are accurately distinguished.

[0061] Figure 4 The diagram illustrates the steps of a method for correcting spatial coordinates. In one embodiment, as shown... Figure 4 As shown, after step 125, the method for controlling personnel in confined spaces in underground engineering also includes: Step 1251: Perform data filtering on the absolute spatial coordinates.

[0062] In this step, considering that dust obstruction and personnel movement may cause positioning noise in a limited space, a Kalman filter algorithm is used to smooth the absolute spatial coordinate data.

[0063] Step 1252: Obtain the fusion vector parameters based on the absolute spatial coordinates and their motion velocity.

[0064] In this step, a fusion vector parameter is calculated based on the absolute spatial coordinates and velocity information. This parameter combines position and velocity information to describe the target's state change trend. This step combines position and velocity information to provide a more comprehensive understanding of the target's dynamic characteristics for subsequent predictions, enhancing the system's ability to understand the target's movement behavior.

[0065] Step 1253: Obtain the current predicted coordinates based on the fusion vector parameters.

[0066] In this step, based on the fusion vector parameters obtained in the previous step, the target's position at the next moment is predicted, thus obtaining the current predicted coordinates. This step, by predicting the target's direction and position in advance, helps to quickly respond to the target's actual movement and supports real-time tracking.

[0067] Step 1254: Obtain the current observed coordinates and correct the current predicted coordinates, and output the corrected absolute spatial coordinates.

[0068] In this step, the actual observed current coordinates are obtained, and then compared and analyzed with the previously predicted coordinates. The differences are used to correct the predicted coordinates, and finally, the corrected absolute spatial coordinates are output. This step, by comparing the predicted values ​​and the actual observed values, can effectively adjust the prediction error, making the system's positioning results closer to the actual situation, and further improving the accuracy and reliability of tracking.

[0069] Figure 5 The diagram illustrates the steps involved in generating a motion trajectory. In one embodiment, as shown... Figure 5 As shown, step 126 includes: Step 1261: Refresh the absolute spatial coordinates based on the preset frequency.

[0070] This step updates the absolute spatial coordinates of personnel entering and exiting the venue at a preset frequency. By periodically refreshing the coordinates, the real-time nature and accuracy of the tracking data can be ensured. It is suitable for dynamic environments, can reflect the movement of the target in a timely manner, and reduce errors caused by delays.

[0071] Step 1262: Connect the continuous absolute spatial coordinates in series according to the timestamp to generate the motion trajectory.

[0072] In this step, the system connects a series of absolute spatial coordinate points obtained in step 1261 according to time sequence to form a line representing the target's movement path, i.e., the motion trajectory. Each coordinate point has a corresponding timestamp, which ensures the correct sorting and association during trajectory construction. The generated motion trajectory provides intuitive data support for analyzing the target's movement patterns. It can not only clearly show the target's route from one place to another, but also be used for further analysis of changes in walking speed, identification of stopping points, etc., which is helpful for understanding the behavioral patterns of analysts in the confined space of underground engineering.

[0073] Furthermore, the preset frequency in step 1261 can be adjusted according to the safety requirement level of the underground engineering space. The baseline value of the preset frequency is set to 500 ms / time. The higher the safety requirement level, the higher the preset frequency. It can be set that the acquisition time corresponding to the preset frequency is halved for each increase in the safety requirement level. Specifically, the preset frequency corresponding to the lowest safety requirement level is 500 ms / time. The preset frequency corresponding to increasing the safety requirement level by one level is 250 ms / time, and the preset frequency corresponding to increasing the safety requirement level by another level is 125 ms / time, and so on.

[0074] In addition, the preset frequency in step 1261 can be adjusted according to a preset adjustment coefficient. The baseline value of the preset frequency is set to 500 ms / time. The higher the adjustment coefficient, the higher the preset frequency. The adjustment coefficient can be set to change proportionally to the preset frequency. Specifically, when the adjustment coefficient is 1, the corresponding preset frequency is 500 ms / time. Increasing the adjustment coefficient by 0.1 reduces the corresponding acquisition time of the preset frequency by 10 ms. For example, when the adjustment coefficient is 1.1, the preset frequency is 490 ms / time, and so on.

[0075] Figure 6 The diagram illustrates the steps of a method for calculating the first correction parameter based on a sliding window. In one embodiment, as shown... Figure 6 As shown, step 170 includes: Step 171: Use the current time as a reference point to backtrack the preset duration setting sliding window.

[0076] In this step, a time sliding window is formed by looking back a fixed period of time (such as 10 minutes, 30 minutes, etc.) from the current time as the endpoint. This window is continuously updated over time to ensure that the data used is timely and representative.

[0077] Step 172: Summarize the total number of people in the sliding window.

[0078] In this step, the number of people passing through the entrance within the sliding window (including net entry and exit flow or total throughput, usually referring to the cumulative number of people passing through in one or two directions) is counted as a raw indicator reflecting the recent activity level of people.

[0079] Step 173: Calculate the average number of people per unit time based on the total number of people and the preset time.

[0080] In this step, the total pedestrian flow is divided by the duration of the sliding window to obtain the average pedestrian flow per unit time, such as "people / minute" or "people / second", which is used to quantify the current pedestrian flow intensity in the underground space.

[0081] Step 174: Set the corresponding benchmark flow rate based on the spatial area of ​​the underground space.

[0082] In this step, based on the actual area of ​​the limited space of the underground project, the capacity of the evacuation passage, safety regulations, etc., a "benchmark reference flow" is pre-set at the entrance and exit of the underground project, such as X people passing through the entrance and exit per 100 square meters per minute or per second as a benchmark for judging whether the flow of people in the underground project is in a normal, busy or peak state.

[0083] Step 175: If the average pedestrian flow is less than or equal to the baseline reference flow, then the first correction parameter is 1.

[0084] Step 176: If the average passenger flow is greater than the baseline reference flow but less than twice the baseline reference flow, then the value range of the first correction parameter is greater than 1 and less than 2.

[0085] Step 177: If the average passenger flow is greater than or equal to twice the baseline reference flow, then the first correction parameter is 2.

[0086] In steps 175 to 177, if the average flow of people is less than or equal to the baseline reference flow, it indicates that the flow of people is stable, and the first correction parameter is 1, that is, the correction of the over-limit working conditions of personnel is not amplified. If the baseline reference flow is less than the average flow and less than 2 × the baseline reference flow, it indicates that the flow of people is relatively busy. The first correction parameter should be greater than 1 and less than 2. The average flow and the first correction parameter can be linearly mapped to each other. If the average passenger flow is ≥2×reference flow, it indicates that the underground project is in a peak state. The first correction parameter = 2 to avoid over-exaggerating the correction of the over-limit working conditions of personnel and thus failing to trigger the evacuation instruction. The mapping relationship in this embodiment enables dynamic threshold adjustment to improve the flexibility of evacuation instructions. The larger the recent flow of people, the higher the upper limit of the number of people allowed to be, but the maximum magnification is limited to avoid failing to trigger the evacuation instruction command.

[0087] Figure 7 The diagram illustrates the steps of adjusting the preset duration of a sliding window. In one embodiment, as shown... Figure 7 As shown, the personnel control method in this underground engineering confined space also includes: Step 710: Set the sensitivity coefficient.

[0088] In this step, the sensitivity coefficient is a configurable parameter (usually a positive real number, such as 0.5, 1.0, 2.0, etc.) used to characterize the system's "response speed" or "smoothness" in response to changes in pedestrian flow. This coefficient can be set according to actual engineering needs: for example, in high-risk operation phases (such as before blasting) where rapid personnel clearance is required, a high sensitivity (i.e., a smaller sliding window) is preferable; while in routine operation phases, a low sensitivity (i.e., a larger window) can be used to suppress noise.

[0089] Step 720: Adjust the preset duration based on the sensitivity coefficient.

[0090] In this step, the preset duration (i.e., the sliding window width) is no longer fixed, but is inversely proportional to the sensitivity coefficient. A typical relationship can be expressed as: T 窗口 =T0 / k; T0 is the baseline window duration, which can be set to a value between 5 minutes and 30 minutes, and k is the sensitivity coefficient; When k>1, the system is in a highly sensitive state. The corrected sliding window width is shorter, which can better focus on recent traffic flow, and the system responds faster and more sensitively. When k < 1, the system is in a low-sensitivity state. The corrected sliding window width becomes longer, paying more attention to long-term traffic flow. The system response is slower but the anti-interference ability is stronger.

[0091] In application, this embodiment can dynamically adapt to different sensitivity coefficient requirements. Under high-risk conditions (such as blasting preparation and gas leak early warning), increasing the sensitivity coefficient and shortening the sliding window width allows the system to quickly capture the completion status of personnel clearing, avoiding delays in issuing blasting commands due to excessive historical data. It effectively optimizes the balance between false alarms and false alarms; high sensitivity reduces false alarms, while low sensitivity reduces false alarms. By adjusting the sensitivity coefficient, the optimal balance between safety and efficiency can be achieved in different scenarios. Adjusting the sensitivity coefficient can change the behavior strategy, supporting remote configuration and automatic switching (such as automatically loading the corresponding sensitivity coefficient according to the process), enhancing engineering applicability.

[0092] Figure 8 The diagram illustrates the steps of a method for correcting over-limit operating conditions in a secondary manner. In one embodiment, as shown... Figure 8 As shown, the personnel control method in this underground engineering confined space also includes: Step 810: Obtain the safety requirement level of the underground space.

[0093] Step 820: Obtain the corresponding second correction factor based on the safety requirement level; the higher the safety requirement level, the smaller the second correction factor.

[0094] Step 830: Perform a second correction on the personnel exceeding the limit conditions based on the second correction coefficient.

[0095] Specifically, the safety requirement levels include general, hazardous, and extremely hazardous levels, with second correction factors of 1, 0.6, and 0.3 corresponding to the general, hazardous, and extremely hazardous levels, respectively.

[0096] In this embodiment, the safety requirement level is a risk level determined by comprehensively considering factors such as the current operation stage of the underground project, environmental risks (such as gas concentration and support status), and process type (such as blasting, tunneling, and formwork).

[0097] This embodiment divides it into three levels: general level (routine operation), hazardous level (potential risk exists), and extremely hazardous level (high-risk operation, such as before blasting, toxic gas leakage, etc.).

[0098] In step 820, a higher safety requirement level indicates a greater risk, and the permitted number of people should be more strictly controlled. Therefore, the second correction factor is inversely proportional to the safety level: General level → Second correction factor = 1 (no reduction in personnel working beyond limits); Hazard level → Second correction factor = 0.6 (personnel exceeding limits reduced to 60%); Extremely dangerous level → Second correction factor = 0.3 (personnel correction conditions are only retained at 30%).

[0099] In step 830, the personnel over-limit working condition is corrected a second time. Based on the personnel over-limit working condition that has been adjusted by the first correction parameter, it is multiplied by the second correction coefficient to obtain the final dynamic personnel over-limit working condition. This process realizes the dual control of "personnel flow adaptability + safety requirement level sensitivity".

[0100] In application, this embodiment implements risk-driven dynamic personnel limit management. In high-risk scenarios, it automatically tightens the personnel limit, forcibly reducing the number of workers on-site, thus reducing the possibility of accidents and injuries at the source and adhering to the safety management principle that the higher the risk, the stricter the control. It effectively improves the compliance and timeliness of emergency response. For example, when entering an "extremely dangerous" state (such as the countdown stage of blasting), the system automatically lowers the over-limit threshold to an extremely low level (e.g., only allowing 3 people). Once exceeded, an evacuation order is triggered to ensure that personnel are minimized before high-risk operations. During normal operation phases, a higher threshold is maintained (the second correction coefficient is 1), without affecting normal construction efficiency. Restrictions are tightened only when necessary, balancing safety and production continuity. This embodiment complements the aforementioned first correction mechanism based on the average flow of people within a sliding window. The first correction parameter addresses fluctuations in flow of people, while the second correction parameter addresses changes in risk level. The combination of these two can construct a more intelligent and reliable personnel capacity control model.

[0101] An example of a personnel control system for confined spaces in underground engineering is as follows: Figure 9 The diagram shown is a schematic representation of a personnel control system for confined spaces in underground engineering, according to an embodiment of this application. This application also provides a personnel control system for confined spaces in underground engineering, such as... Figure 9 As shown, the system includes: a monitoring module 901, a spatial dynamic refresh module 902, an instruction generation module 903, and a correction module 904.

[0102] The monitoring module 901 is configured to obtain entrance monitoring data at the entrance of the underground project based on multiple infrared detection modules.

[0103] The spatial dynamic refresh module 902 is communicatively connected to the monitoring module 901. The spatial dynamic refresh module 902 is configured to: call a preset recognition model to identify personnel entry and exit data based on the entrance monitoring data, and update real-time personnel data based on the personnel entry and exit data; and obtain spatial dynamic data of the internal space of the underground project based on the real-time personnel data.

[0104] The instruction generation module 903 is communicatively connected to the spatial dynamic refresh module 902. The instruction generation module 903 is configured to: generate a blasting instruction if the spatial dynamic data meets the personnel clearing conditions and continues for a first preset duration during the blasting preparation stage; and generate an evacuation instruction if the spatial dynamic data reaches the personnel over-limit condition during the underground operation stage.

[0105] The correction module 904 is communicatively connected to the instruction generation module 903. The correction module 904 is configured to: periodically update the corresponding personnel clearing conditions and personnel over-limit conditions according to different application conditions; set a sliding window based on the current time, calculate the average flow of people in the sliding window, and obtain the corresponding first correction parameter based on the average flow of people; the average flow of people is proportional to the first correction condition; and correct the personnel over-limit condition based on the first correction parameter.

[0106] The basic principles of this application have been described above with reference to specific embodiments. However, it should be noted that the advantages, benefits, and effects mentioned in this application are merely examples and not limitations, and should not be considered as essential features of each embodiment of this application. Furthermore, the specific details disclosed above are for illustrative and facilitative purposes only, and are not limitations. These details do not limit the application to the necessity of employing the aforementioned specific details for implementation.

[0107] The block diagrams of devices, apparatuses, devices, and systems involved in this application are merely illustrative examples and are not intended to require or imply that they must be connected, arranged, or configured in the manner shown in the block diagrams. As those skilled in the art will recognize, these devices, apparatuses, devices, and systems can be connected, arranged, and configured in any manner. Words such as “comprising,” “including,” “having,” etc., are open-ended terms meaning “including but not limited to,” and are used interchangeably with them. The terms “or” and “and” as used herein refer to the terms “and / or,” and are used interchangeably with them unless the context clearly indicates otherwise. The term “such as” as used herein refers to the phrase “such as but not limited to,” and is used interchangeably with it.

[0108] It should also be noted that in the apparatus, equipment, and methods of this application, the components or steps can be disassembled and / or recombined. These disassemblies and / or recombinations should be considered as equivalent solutions of this application.

[0109] The above description of the disclosed aspects is provided to enable any person skilled in the art to make or use this application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the general principles defined herein can be applied to other aspects without departing from the scope of this application. Therefore, this application is not intended to be limited to the aspects shown herein, but rather to be carried out within the widest scope consistent with the principles and novel features of the invention herein.

[0110] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Any modifications or equivalent substitutions made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A method for personnel control in confined spaces of underground engineering projects, characterized in that, include: Entrance monitoring data at the entrance of the underground project was obtained based on multiple infrared detection modules; The system calls a preset recognition model to identify personnel entry and exit data based on the entrance monitoring data, and updates real-time personnel data based on the personnel entry and exit data. Based on the real-time personnel data, the spatial dynamic data of the internal space of the underground project is obtained; During the blasting preparation stage, if the spatial dynamic data matches the personnel clearing conditions and continues for a first preset duration, a blasting instruction is generated. During the underground operation phase, if the spatial dynamic data reaches the point where personnel exceed the limit, an evacuation instruction will be generated. Update the corresponding personnel clearing conditions and personnel over-limit conditions regularly according to different application conditions; Based on the current time, a sliding window is set, the average pedestrian flow within the sliding window is calculated, and the corresponding first correction parameter is obtained based on the average pedestrian flow. The average passenger flow is directly proportional to the first corrected operating condition; The personnel's over-limit working conditions are corrected based on the first correction parameter.

2. The method for personnel control in confined spaces of underground engineering according to claim 1, characterized in that, The entrance monitoring data at the entrance of the underground project obtained based on multiple infrared detection modules includes: Receive individual infrared data detected by each of the infrared detection modules at the entrance of the underground project; The entrance monitoring data is obtained by aggregating multiple individual infrared data.

3. The method for personnel control in confined spaces of underground engineering according to claim 2, characterized in that, The step of calling the preset recognition model to identify personnel entry and exit data based on the entrance monitoring data, and updating real-time personnel data based on the personnel entry and exit data includes: Extract the thermal radiation zone of personnel at 35°C-38°C from the single infrared data; Mark target feature points within the personnel's thermal radiation zone; Based on the module spatial coordinates of the infrared detection module, the relative spatial coordinates between the target feature point and the infrared detection module are located; Generate a relative line connecting the module's spatial coordinates and the relative spatial coordinates, and obtain the corresponding relative spatial angle; The absolute spatial coordinates of the target feature point are calculated based on the multiple sets of relative spatial angles and the multiple module spatial coordinates. By tracking the movement trajectory of the absolute spatial coordinates, the personnel entry and exit data can be obtained; Based on the personnel entry and exit data, if someone enters, the number of people in the venue increases by one; if someone exits, the number of people in the venue decreases by one.

4. The method for personnel control in confined spaces of underground engineering according to claim 3, characterized in that, After calculating the absolute spatial coordinates of the target feature point based on multiple sets of relative spatial angles and multiple module spatial coordinates, the method further includes: Perform data filtering on the absolute spatial coordinates; The fusion vector parameters are obtained based on the absolute spatial coordinates and their motion speed; The current predicted coordinates are obtained based on the fusion vector parameters; Obtain the current observed coordinates, correct the current predicted coordinates, and output the corrected absolute spatial coordinates.

5. The method for personnel control in confined spaces of underground engineering according to claim 4, characterized in that, The process of tracking the motion trajectory of the absolute spatial coordinates to obtain the personnel entry and exit data includes: The absolute spatial coordinates are refreshed based on a preset frequency; The motion trajectory is generated by connecting the consecutive absolute spatial coordinates according to timestamps.

6. The method for personnel control in confined spaces of underground engineering according to claim 1, characterized in that, The step of setting a sliding window based on the current time, calculating the average pedestrian flow within the sliding window, and obtaining the corresponding first correction parameter based on the average pedestrian flow includes: The sliding window is set back by a preset time period using the current time as a reference point; Summarize the total pedestrian flow within the sliding window; Calculate the average number of people per unit time based on the total number of people and the preset time. A corresponding baseline reference flow rate is set based on the spatial area of ​​the underground space; If the average pedestrian flow is less than or equal to the baseline reference flow, then the first correction parameter is 1; If the average passenger flow is greater than the benchmark reference flow but less than twice the benchmark reference flow, then the value range of the first correction parameter is greater than 1 and less than 2. If the average pedestrian flow is greater than or equal to twice the baseline reference flow, then the first correction parameter is 2.

7. The method for personnel control in confined spaces of underground engineering according to claim 6, characterized in that, Also includes: Set the sensitivity coefficient; The preset duration is adjusted based on the sensitivity coefficient.

8. The method for personnel control in confined spaces of underground engineering according to claim 1, characterized in that, Also includes: Obtain the safety requirement level of the underground space; The corresponding second correction factor is obtained based on the aforementioned safety requirement level; The higher the safety requirement level, the smaller the second correction factor; The personnel over-limit working conditions are corrected a second time based on the second correction coefficient.

9. The method for personnel control in confined spaces of underground engineering according to claim 8, characterized in that, The safety requirement levels include a general level, a dangerous level, and an extremely dangerous level, and the second correction coefficients corresponding to the general level, the dangerous level, and the extremely dangerous level are 1, 0.6, and 0.3, respectively.

10. A personnel control system for confined spaces in underground engineering, characterized in that, include: The monitoring module is configured to obtain entrance monitoring data at the entrance of the underground project based on multiple infrared detection modules; A spatial dynamic refresh module is communicatively connected to the monitoring module. The spatial dynamic refresh module is configured to: call a preset recognition model to identify personnel entry and exit data based on the entrance monitoring data, and update real-time personnel data based on the personnel entry and exit data; and obtain spatial dynamic data of the internal space of the underground project based on the real-time personnel data. The instruction generation module is communicatively connected to the spatial dynamic refresh module. The instruction generation module is configured to: generate a blasting instruction if the spatial dynamic data meets the personnel clearing conditions and continues for a first preset duration during the blasting preparation stage; and generate an evacuation instruction if the spatial dynamic data reaches the personnel over-limit condition during the underground operation stage. The correction module is communicatively connected to the instruction generation module. The correction module is configured to periodically update the corresponding personnel clearing conditions and personnel over-limit conditions according to different application conditions. Based on the current time, a sliding window is set, the average pedestrian flow within the sliding window is calculated, and the corresponding first correction parameter is obtained based on the average pedestrian flow. The average passenger flow is proportional to the first corrected working condition; the personnel over-limit working condition is corrected based on the first corrected parameter.

Images