Intelligent management method and platform based on construction safety penetration type management

Abstract

The invention provides an intelligent management method and platform based on construction safety penetration type management. The method comprises the following steps: constructing a construction safety state structural body of a construction site; designing a task priority scoring function to calculate scheduling priority scores of the tasks, and generating a security task set based on the scheduling priority scores; obtaining a recommended candidate person set, sorting the final scheduling scores according to the scheduling priority scores, selecting the candidate person with the maximum final scheduling score score to assign the task, and performing final scheduling score calculation on all candidate persons to obtain a scheduling output structure; the recommended candidate person and member set comprises a plurality of candidate persons and members; and issuing a task instruction according to the scheduling output structure, and calculating the task completion degree based on the multi-source verification information to confirm whether the task is actually executed and closed in a compliance manner. A full-chain management system is formed, and penetrating type landing of construction safety management is truly achieved.

Description

Technical Field

[0001] This invention belongs to the field of construction safety penetration management, and particularly relates to intelligent management methods and platforms based on construction safety penetration management. Background Technology

[0002] With the continuous expansion of construction projects and the increasing complexity of construction sites, safety management issues are becoming increasingly prominent. Traditional construction safety management methods mainly rely on manual inspections, periodic reports, and post-event corrections, a model that suffers from significant lag and uncertainty. Although video surveillance, positioning systems, and IoT sensors have been introduced in recent years, these methods are mostly used in isolation, failing to create a holistic and dynamic understanding of construction safety, and even more so, making it difficult to effectively transmit management objectives to the front-line operational level. Managers often perceive safety risks but lack a mechanism that can penetrate levels and reach directly to the executors, leading to a disconnect between task instructions and actual execution, creating a situation where "risks are visible, but actions cannot be controlled."

[0003] On the other hand, the construction environment is highly dynamic and uncertain, with frequent personnel movement, tasks being adjusted at any time, and risk sources constantly changing. If the management system cannot synchronize this dynamic state in real time and generate clear execution paths, it will cause scheduling chaos and execution gaps.

[0004] At the same time, even if tasks have been assigned, there is often a lack of effective confirmation mechanisms to ensure that tasks are indeed completed or completed according to standards, which directly affects the effectiveness of safety closed-loop management.

[0005] Therefore, existing technologies are still at the stage of data collection and static analysis, lacking a holistic method that can dynamically transform safety strategies into tasks and implement them, making it difficult to meet the complex safety management needs of multiple roles, multiple levels, and multiple scenarios on construction sites. Summary of the Invention

[0006] The purpose of this invention is to propose an intelligent management method and platform based on the penetrating management of construction safety, so as to solve the above-mentioned problems.

[0007] To achieve the above objectives, a first aspect of the present invention provides an intelligent management method based on construction safety penetration management, the method comprising the following steps: S1. Construct a construction safety status structure for the construction site; wherein, the construction safety status structure includes the operational status of the construction area, the real-time three-dimensional coordinates of the workers, the risk score of the area, and the result of whether the workers are performing the corresponding tasks specified in the task plan in the area. If yes, it is 1; otherwise, it is 0. S2. Design a task priority scoring function to calculate the scheduling priority score of the task, and generate a set of safe tasks based on the scheduling priority score; S3. Obtain the recommended candidate set, calculate the comprehensive scheduling matching score by combining the safety task set and the construction safety status structure, and correct it based on the scheduling cost penalty item to obtain the final scheduling score. Sort the final scheduling scores according to the scheduling priority scores, select the candidate with the highest final scheduling score to assign the task, calculate the final scheduling score for all the candidates, and obtain the scheduling output structure; the recommended candidate set contains several candidates. S4. Issue task instructions according to the scheduling output structure, and calculate the task completion rate based on multi-source verification information to confirm whether the task has been actually executed and closed in compliance with regulations.

[0008] Furthermore, the risk score for the area is generated by a weighted average calculation based on personnel density and operational risk level; The result of whether the personnel are performing the tasks specified in their task plan in the area is generated by comparing the real-time three-dimensional coordinates of the personnel with their task assignment field. If their location matches the task area and the current time is within the task plan time period, the result is 1; otherwise, it is 0.

[0009] Furthermore, S2 specifically includes: The actual on-duty personnel ratio in the current construction area is obtained and matched with the task allocation ratio. Combined with the risk score of the area, and introducing a task scheduling difficulty regularization term, a priority scoring function is designed to obtain the scheduling priority score of the corresponding task. The task scheduling difficulty regularization term is used to penalize those tasks that are high-risk but difficult to schedule. By combining task type, target area number, candidate personnel, task time window, area risk score, scheduling priority score, and task scheduling difficulty regularization term, a regional construction task structure and a set of security tasks are constructed.

[0010] Furthermore, the task scheduling difficulty regularization term is generated by weighted summation based on the complexity of the physical resources required by the current task, the average spatial distance between the current candidate personnel and the task target area, and the degree of conflict between the current task and other tasks in terms of personnel, time, and space.

[0011] Furthermore, the comprehensive scheduling matching score is calculated and generated based on the distance of the candidate to the region, whether the candidate performs the tasks specified in the task plan in the region, the candidate's task load index, and the candidate's historical performance score. The historical performance score is generated based on the individual's task completion rate and violation records over the past 7 days, with a value range of... .

[0012] Further, the step of sorting the candidates based on the final scheduling scores according to the scheduling priority scores and selecting the candidate with the highest final scheduling score to assign the task specifically includes: Obtain all the comprehensive scheduling matching scores, and adopt a task-first greedy scheduling mechanism. In each round, the task with the highest scheduling priority score is selected first, and then the candidate with the highest comprehensive scheduling matching score is selected from its candidate set to assign the task. Record the responsibility mapping between the task and the candidate. Each candidate can only be assigned one task, and after being assigned, it will not participate in other task candidate sets until the current round of scheduling is completed.

[0013] Furthermore, the final scheduling score is generated based on a scheduling cost penalty term, and scores are corrected for tasks with excessive spatial movement costs or severe scheduling conflicts.

[0014] Furthermore, S4 specifically includes: Each candidate can only be assigned one task, and the candidate is designated as the person in charge. The system sends interactive instruction packets to the responsible person based on the task type; The interactive instruction package is executed, feedback information is collected, a task risk compensation item is introduced, a task completion function is constructed, and the task completion degree is calculated; wherein, the feedback information includes location, image, and device feedback. If the task completion rate does not reach the completion threshold, it is marked as incomplete, and a prompt is automatically generated and fed back to the scheduling platform for reassignment or review. If completed, a completion structure will be output; wherein the completion structure includes task number, responsible person ID, execution time window, each score item, task completion degree, task completion label, and failure reason suggestion fields.

[0015] Furthermore, the task completion score is calculated based on the location matching score, behavior recognition score, subjective confirmation score, and task risk compensation item. The location matching score indicates whether the person in charge appears in the area within the task time window; the behavior recognition score is the confidence level that the current frame image belongs to the task behavior; the subjective confirmation score is obtained by the worker actively scanning the code or clicking the completion confirmation button on the terminal; and the task risk compensation item indicates the compensation for misjudgment caused by the operational risk level of the task.

[0016] In a second aspect, the present invention provides an intelligent management platform based on construction safety penetration management, the platform comprising: The state modeling unit is used to construct the construction safety state structure of the construction site; wherein, the construction safety state structure includes the operation status of the construction area, the real-time three-dimensional coordinates of the workers, the risk score of the area, and the result of whether the workers are performing the tasks specified in their task plan in the area, which is 1 if yes and 0 otherwise. The task generation unit is used to design a task priority scoring function to calculate the scheduling priority score of a task, and generate a set of safe tasks based on the scheduling priority score. The task scheduling unit is used to obtain a set of recommended candidates, calculate a comprehensive scheduling matching score by combining the set of safety tasks and the construction safety status structure, and make corrections based on the scheduling cost penalty to obtain a final scheduling score. The final scheduling score is then sorted according to the scheduling priority score, and the candidate with the highest final scheduling score is selected to be assigned the task. The final scheduling score is calculated for all candidates to obtain a scheduling output structure. The set of recommended candidates contains several candidates. The execution verification unit is used to issue task instructions according to the scheduling output structure and calculate the task completion degree based on multi-source verification information to confirm whether the task has been actually executed and closed in compliance with regulations.

[0017] The beneficial technical effects of the present invention are at least as follows: This invention addresses the challenges in construction site safety management, such as the difficulty in penetrating management intent to the execution level, the lack of dynamic adaptation in task scheduling, and the difficulty in confirming task execution results. It proposes a penetrating management method and platform based on state modeling, task generation, scheduling optimization, and execution verification. By establishing a construction state structure with work status, personnel location, and risk level as core fields, a structured expression of the dynamic state of the site is achieved. A task set is generated based on state gaps, and a scheduling feasibility regularization term is introduced, ensuring that tasks not only possess the semantics of "what should be done" but also structured information on "whether it can be done." In the scheduling phase, a multi-factor scheduling model integrating spatial distance, personnel load, historical performance reliability, and conflict costs is designed, making task allocation more aligned with site resource conditions and ensuring executability. In the task execution phase, a multi-source verified task completion function is constructed, and a risk compensation regularization term is innovatively introduced to improve the accuracy of judgment in high-risk work scenarios, ensuring that tasks are actually executed and properly closed. This invention, through the organic connection of multiple steps, forms a full-chain management system from strategy to execution, from task to action, and from action to confirmation, truly achieving penetrating implementation of construction safety management. Attached Figure Description

[0018] The present invention will be further described with reference to the accompanying drawings, but the embodiments in the drawings do not constitute any limitation on the present invention. For those skilled in the art, other drawings can be obtained based on the following drawings without creative effort.

[0019] Figure 1 This is a flowchart of the intelligent management method based on the penetrating management of construction safety of the present invention. Detailed Implementation

[0020] Embodiments of the present invention are described in detail below. Examples of these embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0021] like Figure 1 As shown in the embodiment of the present invention, an intelligent management method based on penetrating management of construction safety is provided, the method comprising: S1. Construct a construction safety status structure for the construction site; wherein, the construction safety status structure includes the operational status of the construction area, the real-time three-dimensional coordinates of the workers, the risk score of the area, and the result of whether the workers are performing the tasks specified in their task plan in the area. If yes, it is 1, otherwise it is 0.

[0022] Specifically, this step aims to establish a construction site status representation structure oriented towards task-driven and safety management. This structure is not a generalized state model, but rather a specially constructed computable state semantic expression framework that serves the subsequent security task generation and scheduling mechanism.

[0023] Furthermore, the input data used in this step includes: work plan data from the construction management system, recorded as a triplet structure {area number, work type, start and end time}, which is uniformly converted by the system into a structured format as work plan fields; and personnel location information obtained through the UWB positioning system, recording the three-dimensional coordinates of each worker once per second, and identifying them as... The system extracts risk marker fields by taking the high-risk operation registration data submitted daily by the construction site safety officer and extracting the operation type and time range; and the system also uses the construction area division data initially set for spatial boundary determination.

[0024] After the data preparation is complete, the system uses the current time. For indexing, construct the following construction safety status structure. :

[0025] in, Indicates the number is The construction area is constantly The work status is categorized into "Under Construction," "Not Started," and "Unplanned Work," among others. This field is obtained by comparing the current time with the start and end times in the work plan, and is forcibly marked as under supervision when there are high-risk temporary operations. It is the first The real-time 3D coordinates of each worker are collected in real time through the positioning system and updated once per second, which is used for spatial judgment and calculation of personnel distribution density.

[0026] Indicates the area At any moment The risk score, calculated by weighting personnel density and operational risk level, is expressed as:

[0027] in, It is a region Inner Moments The number of identified workers was determined by comparison. Derived from the coordinates of the region boundary; Let be a Boolean conditional function, when the first... Personnel are in the area Furthermore, the value is 1 when the planned work location matches the actual location, and 0 otherwise; It is the first The risk coefficient corresponding to the type of work currently performed by each person is taken from a predefined work type and risk level mapping table, such as 0.9 for high-altitude work, 0.8 for welding work, and 0.3 for general material handling.

[0028] In actual deployment, for example, in area A5, a total of 5 workers were identified, 3 of whom were performing electric welding operations. ), 2 people were performing the handling work ( If the risk score for this area is:

[0029] Personnel Is it in the region? The task is carried out according to the task plan. This value is determined by comparing personnel positioning. If the job assignment field is generated in a way that matches the job area and the current time is within the task's scheduled time period, then... Otherwise, it is 0. This field is used in subsequent steps to determine the task execution status and task scheduling compatibility.

[0030] S2. Design a task priority scoring function to calculate the scheduling priority score of the task, and generate a set of safe tasks based on the scheduling priority score.

[0031] Specifically, this step is based on the construction safety status structure constructed in the previous step. This is used to generate a set of task structures that have scheduling feasibility and awareness of actual field constraints. These tasks are not only "safety behaviors that should be completed," but also task units with scheduling attributes in their structure. Unlike traditional safety inspection tasks, these tasks are generated entirely based on the current state rather than a preset template, and the task structure embeds scheduling auxiliary information such as personnel matching degree, job accessibility, and resource requirement complexity, thus laying an operational decision-making foundation for scheduling optimization of subsequent steps.

[0032] The core idea behind task generation is to identify "state gaps," that is, state models. The implicit situations of "actions that should have been performed but have not been performed" manifest in gaps such as: lack of supervision in high-risk areas, absence of on-site personnel, and lack of isolation warnings in overlapping work scenarios. These gaps are identified from the status field. Obtain the current operation status of each region. Reflecting the risk intensity of the area, It is the real-time spatial location of personnel, and This indicates whether personnel have appeared in the corresponding area as planned.

[0033] Furthermore, the system first filters out all regions that meet the following conditions. The current work status is under construction or under enhanced supervision (i.e., ∈{“Under Construction”, “Enhanced Supervision”}), and risk score Higher than the preset threshold These types of areas require priority task generation. Then, check the task matching for personnel in that area; if there is a match... The proportion of personnel exceeds This indicates that there is a gap in execution capacity or an operation that exceeds the scope of the current area, triggering the task generation logic.

[0034] Furthermore, to generate more meaningful task structures for scheduling, we introduce a task priority scoring function for "penetrating scheduling." This function considers not only regional risk and personnel availability but also models "task schedulability." This is represented by a regularization term describing the task's suitability to current resource conditions. The priority scoring function is defined as follows:

[0035] in, For the task Scheduling priority score, This is the area where the task belongs. Risk score; This indicates the proportion of personnel actually on duty and matched with task assignments in the region, reflecting the current level of vacancies. It is a task scheduling difficulty regularization term, used to penalize tasks that are high-risk but difficult to schedule. The weighting parameters are adjustable and are fixed after being adjusted based on historical data in actual system deployment.

[0036] The construction of this step is a key innovation, reflecting the "scheduling feasibility" of the task from the perspective of system allocation, and is composed of three weighted sub-factors:

[0037] in, This indicates the complexity of the physical resources required for the task, mapped by the task type (e.g., setting up a fence requires a toolbox and two people working together, which is 0.8; posting a warning sign requires only one person manually, which is 0.2). The system determines the average spatial distance between the currently recommended executor and the target area of ​​the task based on... Calculate all available personnel to the region The average distance; To determine the degree of conflict between this task and other tasks in terms of personnel, time, and space, the system uses historical statistical overlap data generated from the previous 5 minutes of tasks for calculation. Each of the three factors is multiplied by a weight. The system is designed by systems engineering to ensure that task generation has dynamic scheduling feasibility assessment capabilities.

[0038] Furthermore, the system constructs tasks for each region that meets the conditions. Each task structure contains the following fields: task type (e.g., setting up a fence, cross-operation identification), and target area number. Recommended candidate set of executors (through comparison) and Confirm), task time window (based on) The corresponding work plan window is automatically generated), and the task risk level (i.e. Task scheduling priority scoring and scheduling difficulty items .

[0039] For example, if in At what time, the state of region A7 (Under construction) There are 6 people in the area, and the task matching rate is... Meanwhile, the current system calculates its scheduling difficulty item. In setting parameters , , At that time, we can obtain:

[0040] The system determines that this task has a high priority and writes it to... The task set includes all structure fields for subsequent scheduling steps.

[0041] The final output is a set of security tasks. Each of them All of them are task units with well-defined structures and rich data fields, and they already have the necessary attributes for direct scheduling and execution. This not only solves the question of "what to do", but also partially answers the questions of "who does it, when to do it, and whether it can be done".

[0042] S3. Obtain the recommended candidate set, calculate the comprehensive scheduling matching score by combining the safety task set and the construction safety status structure, and correct it based on the scheduling cost penalty item to obtain the final scheduling score. Sort the final scheduling scores according to the scheduling priority scores, select the candidate with the highest final scheduling score to assign the task, calculate the final scheduling score for all the candidates, and obtain the scheduling output structure; the recommended candidate set contains several candidates.

[0043] Specifically, this step is part of the security task set. Based on the existing structure and its condition at the construction site Implement a safety task resource scheduling and responsibility allocation mechanism oriented towards construction scenario constraints, and output a scheduling plan. The core innovation of this step lies in the fact that traditional scheduling schemes are mainly based on time or idle time for matching, while this invention, based on the special needs of construction safety management, combines multiple practical factors such as task risk level, resource scheduling cost, personnel execution reliability, and space scheduling cost, and constructs scheduling priority relationships with task penetration and executability as the core optimization goal, thereby realizing a scheduling system with scene awareness capabilities.

[0044] Furthermore, the input to the scheduling mechanism is determined by the state and task structure provided in the first two steps. Among these, Represents the first in the task set A task item includes the fields: task type and target region. Task risk assessment Recommended Candidate Set Scheduling score item Task priority scoring State structure Provide the current location of the workers Personnel task allocation status Operation status These structural fields provide a data foundation for feasibility assessment and cost modeling of scheduling.

[0045] In actual construction, scheduling faces the following limitations: personnel may be located far from the task area, causing response delays; the same person cannot complete multiple physically conflicting tasks in a short period (e.g., insufficient interval between operations in area A and area B); some high-risk tasks need to be assigned to qualified personnel with low workloads; and some tasks cannot be executed immediately due to material constraints. Therefore, the system scheduling goal is not only to "assign and complete all tasks," but also to construct a "penetrating execution" responsibility mapping. To achieve this goal, the scheduling process is designed as follows.

[0046] Furthermore, the system first calculates the task. With its candidates Comprehensive scheduling matching score The scoring function integrates spatial distance, current load, skill matching degree, risk compensation requirements, and historical performance reliability, and is expressed as follows:

[0047] in, Indicates candidates To the area The distance, through Calculated from the coordinates of the geometric center of the region, and normalized to ; The result is whether the personnel defined in step one are performing the tasks specified in their task plan within the area; Personnel The task load metric is calculated by weighting the number of tasks performed in the past 20 minutes. Personnel Historical performance score is generated by the system based on fields such as the individual's task completion rate and violation records over the past 7 days, with a value range of... This is used to provide a compensation mechanism for situations where high mission reliability is required. to The weights set for the system are configured by default. Parameters can be adjusted according to the project's safety management strategy.

[0048] The innovation of the above scoring model lies in the introduction of... Performance scoring, unlike the "experience level" commonly found in traditional scheduling, is dynamically calculated and generated after the reliability of historical behavior is recorded in a structured manner. It can effectively solve the scheduling uncertainty caused by frequent changes in on-site personnel and asymmetric experience.

[0049] Subsequently, the system will... By task priority The task is sorted using a task-first greedy scheduling mechanism, with priority selection made in each round. The biggest task is to select from its pool of candidates. The highest-ranking personnel assign the task, recording the responsibility mapping between the task and the personnel. Each personnel can only be assigned one task, and after assignment, they will not participate in other task candidate sets until the current round of scheduling is completed.

[0050] To further improve scheduling controllability, the system introduces a scheduling cost penalty term during the matching process. Tasks with excessive spatial movement costs or severe scheduling conflicts are scored and corrected to generate a final scheduling score. The definition is as follows:

[0051] in, For the task Assigned to personnel The anticipated resource conflict costs that follow consist of two parts: first, if the person participates in other task candidate sets simultaneously, which tasks will lose available scheduling candidates, and the system will count the number of lost tasks; second, the personnel... If the current location is in a remote area and the expected response time of the task exceeds the threshold, the system will impose a delay penalty. This is the conflict cost weighting coefficient, used to limit unreasonable scheduling paths.

[0052] For example, task This falls under hoisting guidance; risk assessment. Task priority Candidate concentration It has the highest score, but it is currently in area A1, more than 100 meters away from the target area A8, and is also on another high-priority task. System evaluation in the candidate set Therefore, its final score is:

[0053] like The original score is slightly lower, but the conflict cost is 0.1, then The system will eventually assign the task to .

[0054] Final system output scheduling output structure each Fields included: Task ID, Task Content, Responsible Person ID, Target Area Number, Execution Time Window, Scheduling Cost Score The scheduling rationale (such as "high task priority + proximity in space + no resource conflicts") will be used for subsequent execution instruction triggering and control modules.

[0055] S4. Issue task instructions according to the scheduling output structure, and calculate the task completion rate based on multi-source verification information to confirm whether the task has been actually executed and closed in compliance with regulations.

[0056] Specifically, this step is based on the scheduling output structure. A task execution control and verification mechanism is constructed, consisting of "command issuance—action triggering—state observation—structure closure," and the output is an execution control and completion record structure. This step not only achieves the "instruction delivery" of the task, but also ensures that management instructions penetrate from the top level to the front-line construction behavior by designing a structured "behavioral completion modeling method". It can be verified, recorded and closed-loop in the system, thus completing the core logic of the entire construction safety management: whether the task is actually executed, who executes it, and whether it is completed in compliance with regulations.

[0057] Furthermore, the execution verification is divided into two stages: task activation instruction issuance and structured confirmation of task completion. In the first stage, the system sends an interactive instruction package to the responsible person (candidate) according to the task type, including text description, target location map, task time limit, hazard warning, tool list, etc. The push format includes mobile terminal messages, voice reminders, and vibration reminders from wearable smart badges to ensure that the task notification is implemented.

[0058] The second stage is the verification and modeling of task execution behavior. Its core is to construct a task completion function through multi-dimensional sensing methods (feedback information includes localization, images, and device feedback). Furthermore, it embeds a task risk compensation feature specifically designed for construction scenarios. This feature is used to structurally adjust high-risk or visually misjudgment-prone tasks, enabling the verification mechanism to possess both practical differentiation and error control capabilities. The definition is as follows:

[0059] in, Rate the task completion rate. To determine the matching score, the responsible party is indicated. Does the character appear in the area during the mission time window? ,Depend on The overlap rate with the geometric boundary of the region is calculated as follows: 1 for complete overlap and 0 for non-entry. To score the behavior recognition, images from fixed-point cameras deployed in the task area are used as input to a lightweight image recognition model (e.g., MobileNetV3) to classify the target behavior, such as "escalator positioning" or "hanging on the fence". The model outputs the confidence score of the frame image belonging to the task behavior. Finally, the recognition results of multiple frames are averaged by sliding window within the task time window. The score is subjectively confirmed and obtained by the worker actively scanning the code or clicking the confirmation button on the terminal; This is a task risk compensation item, representing compensation for misjudgments caused by the operational risk level of the task, preventing high-risk operations from being mistakenly judged as "incomplete" due to unclear actions.

[0060] The innovative design lies in the fact that traditional systems are prone to misjudging "completed but judged as failed" in high-risk operations due to visual obstruction and personnel protective clothing concealing movements. This mechanism utilizes the steps in S1... The compensation factor is constructed by mapping the regional risk score to the task type in step two, and is expressed as follows:

[0061] in Real-time risk assessment for the area where the mission is located; The visual ambiguity coefficient for task type is set by safety management personnel or learned by the system. For example, "wearing a seat belt" is 0.8 and "cleaning scaffolding" is 0.4, which represents the probability of ambiguity judgment of the task at the visual end. To compensate for the proportional weighting, default values ​​are set during system deployment, such as... .

[0062] For example, if the task For "installing edge fencing", area risk scoring visual blur of the task If the average confidence level of behavior recognition is 0.81, the location matching is 1, and no QR code confirmation is received, then:

[0063] If the scores for each item are , , The final task completion rate is:

[0064] If the system determines that the task has not reached the completion threshold (e.g., set to 0.75), it will mark it as "incomplete" and automatically generate a prompt to be sent to the scheduling platform for reassignment or review.

[0065] The system finally outputs the completed structure. Each structure includes a task number, responsible person ID, execution time window, various scoring items, and task completion rate. The system includes fields such as task completion tags (completed / incomplete / abnormal) and failure reason suggestions (e.g., "high ambiguity of behavior + no confirmation"), which support subsequent task closed-loop statistics, performance data archiving, and automatic accountability chain tracing.

[0066] In another embodiment of the present invention, an intelligent management platform based on construction safety penetration management is provided, the platform comprising: The state modeling unit is used to construct the construction safety state structure of the construction site; wherein, the construction safety state structure includes the operation status of the construction area, the real-time three-dimensional coordinates of the workers, the risk score of the area, and the result of whether the workers are performing the tasks specified in their task plan in the area, which is 1 if yes and 0 otherwise. The task generation unit is used to design a task priority scoring function to calculate the scheduling priority score of a task, and generate a set of safe tasks based on the scheduling priority score. The task scheduling unit is used to obtain a set of recommended candidates, calculate a comprehensive scheduling matching score by combining the set of safety tasks and the construction safety status structure, and make corrections based on the scheduling cost penalty to obtain a final scheduling score. The final scheduling score is then sorted according to the scheduling priority score, and the candidate with the highest final scheduling score is selected to be assigned the task. The final scheduling score is calculated for all candidates to obtain a scheduling output structure. The set of recommended candidates contains several candidates. The execution verification unit is used to issue task instructions according to the scheduling output structure and calculate the task completion degree based on multi-source verification information to confirm whether the task has been actually executed and closed in compliance with regulations.

[0067] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0068] In the embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection of apparatuses or units may be electrical, mechanical, or other forms.

[0069] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0070] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. An intelligent management method based on penetrating management of construction safety, characterized in that: The method includes the following steps: S1. Construct a construction safety status structure for the construction site; wherein, the construction safety status structure includes the operational status of the construction area, the real-time three-dimensional coordinates of the workers, the risk score of the area, and the result of whether the workers are performing the corresponding tasks specified in the task plan in the area. If yes, it is 1; otherwise, it is 0. S2. Design a task priority scoring function to calculate the scheduling priority score of the task, and generate a set of safe tasks based on the scheduling priority score; S3. Obtain the recommended candidate set, calculate the comprehensive scheduling matching score by combining the safety task set and the construction safety status structure, and correct it based on the scheduling cost penalty item to obtain the final scheduling score. Sort the final scheduling scores according to the scheduling priority scores, select the candidate with the highest final scheduling score to assign the task, calculate the final scheduling score for all the candidates, and obtain the scheduling output structure; the recommended candidate set contains several candidates. S4. Issue task instructions according to the scheduling output structure, and calculate the task completion rate based on multi-source verification information to confirm whether the task has been actually executed and closed in compliance with regulations.

2. The intelligent management method based on construction safety penetration management according to claim 1, characterized in that, The risk score for the area is generated by a weighted average of personnel density and operational risk level. The result of whether the personnel are performing the tasks specified in their task plan in the area is generated by comparing the real-time three-dimensional coordinates of the personnel with their task assignment field. If their location matches the task area and the current time is within the task plan time period, the result is 1; otherwise, it is 0.

3. The intelligent management method based on construction safety penetration management according to claim 1, characterized in that, S2 specifically includes: The actual on-duty personnel ratio in the current construction area is obtained and matched with the task allocation ratio. Combined with the risk score of the area, and introducing a task scheduling difficulty regularization term, a priority scoring function is designed to obtain the scheduling priority score of the corresponding task. The task scheduling difficulty regularization term is used to penalize those tasks that are high-risk but difficult to schedule. By combining task type, target area number, candidate personnel, task time window, area risk score, scheduling priority score, and task scheduling difficulty regularization term, a regional construction task structure and a set of security tasks are constructed.

4. The intelligent management method based on construction safety penetration management according to claim 3, characterized in that, The task scheduling difficulty regularization term is generated by weighted summation based on the complexity of the physical resources required by the current task, the average spatial distance between the current candidate personnel and the task target area, and the degree of conflict between the current task and other tasks in terms of personnel, time, and space.

5. The intelligent management method based on construction safety penetration management according to claim 1, characterized in that, The comprehensive scheduling and matching score is calculated and generated based on the distance of the candidate to the region, whether the candidate performs the tasks specified in the task plan in the region, the candidate's task load index, and the candidate's historical performance score. The historical performance score is generated based on the individual's task completion rate and violation records over the past 7 days, with a value range of... .

6. The intelligent management method based on construction safety penetration management according to claim 5, characterized in that, The step of sorting candidates by scheduling priority based on their final scheduling scores and selecting the candidate with the highest final scheduling score to assign the task specifically includes: Obtain all the comprehensive scheduling matching scores, and adopt a task-first greedy scheduling mechanism. In each round, the task with the highest scheduling priority score is selected first, and then the candidate with the highest comprehensive scheduling matching score is selected from its candidate set to assign the task. Record the responsibility mapping between the task and the candidate. Each candidate can only be assigned one task, and after being assigned, it will not participate in other task candidate sets until the current round of scheduling is completed.

7. The intelligent management method based on construction safety penetration management according to claim 5, characterized in that, The final scheduling score is generated based on the scheduling cost penalty, and the score is corrected for tasks with excessive spatial movement costs or serious scheduling conflicts.

8. The intelligent management method based on construction safety penetration management according to claim 1, characterized in that, S4 specifically includes: Each candidate can only be assigned one task, and the candidate is designated as the person in charge. The system sends interactive instruction packets to the responsible person based on the task type; The interactive instruction package is executed, feedback information is collected, a task risk compensation item is introduced, a task completion function is constructed, and the task completion degree is calculated; wherein, the feedback information includes location, image, and device feedback. If the task completion rate does not reach the completion threshold, it is marked as incomplete, and a prompt is automatically generated and fed back to the scheduling platform for reassignment or review. If completed, a completion structure will be output; wherein the completion structure includes task number, responsible person ID, execution time window, each score item, task completion degree, task completion label, and failure reason suggestion fields.

9. The intelligent management method based on construction safety penetration management according to claim 8, characterized in that, The task completion score is calculated based on the location matching score, behavior recognition score, subjective confirmation score, and task risk compensation item. The location matching score indicates whether the person in charge appears in the area within the task time window; the behavior recognition score is the confidence level that the current frame image belongs to the task behavior; the subjective confirmation score is obtained by the worker actively scanning the code or clicking the completion confirmation button on the terminal; and the task risk compensation item indicates the compensation for misjudgment caused by the operational risk level of the task.

10. An intelligent management platform based on penetrating management of construction safety, characterized in that: The platform includes: The state modeling unit is used to construct the construction safety state structure of the construction site; wherein, the construction safety state structure includes the operation status of the construction area, the real-time three-dimensional coordinates of the workers, the risk score of the area, and the result of whether the workers are performing the tasks specified in their task plan in the area, which is 1 if yes and 0 otherwise. The task generation unit is used to design a task priority scoring function to calculate the scheduling priority score of a task, and generate a set of safe tasks based on the scheduling priority score. The task scheduling unit is used to obtain a set of recommended candidates, calculate a comprehensive scheduling matching score by combining the set of safety tasks and the construction safety status structure, and make corrections based on the scheduling cost penalty to obtain a final scheduling score. The final scheduling score is then sorted according to the scheduling priority score, and the candidate with the highest final scheduling score is selected to be assigned the task. The final scheduling score is calculated for all candidates to obtain a scheduling output structure. The set of recommended candidates contains several candidates. The execution verification unit is used to issue task instructions according to the scheduling output structure and calculate the task completion degree based on multi-source verification information to confirm whether the task has been actually executed and closed in compliance with regulations.

Images