A construction project safety management system

Abstract

The application provides an engineering project safety management system, which comprises the following steps: a first calculation unit acquires equipment state indexes, environmental change indexes and personnel behavior indexes, performs weighted calculation, and obtains a comprehensive safety state result. An evaluation unit performs risk evaluation according to the comprehensive safety state result and the personnel behavior indexes, obtains a safety risk evaluation result, and maps the safety risk evaluation result to a risk level identifier according to a preset multi-level boundary threshold. A second calculation unit acquires a basic action intensity coefficient corresponding to the risk level identifier and a reference upper boundary value, performs action intensity calculation according to the basic action intensity coefficient, the reference upper boundary value and the safety risk evaluation result, and obtains an action intensity value. A management unit generates a management instruction set according to the risk level identifier, the action intensity value and a preset template level table, and performs safety management control on an engineering project site according to the management instruction set. The accuracy of engineering project risk identification and the safety of the engineering project are improved.

Description

Technical Field

[0001] This invention belongs to the field of management systems, and particularly relates to a safety management system for engineering projects. Background Technology

[0002] As engineering projects become increasingly large-scale, specialized, and multi-process integrated, safety risks on construction sites are no longer limited to single equipment malfunctions or individual violations. Instead, they increasingly manifest as a complex risk arising from the simultaneous superposition and amplification of abnormal equipment operation, changing environmental conditions, and unsafe personnel behaviors. In construction scenarios such as building buildings, municipal utilities, deep foundation pits, and high-altitude hoisting, on-site risks exhibit significant dynamism, time-varying characteristics, and regional variations. The degree of danger of the same type of safety hazard varies across different construction stages, environmental conditions, and work areas. Current engineering project safety management technologies largely rely on manual inspections, scheduled checks, post-event rectification, or single-item monitoring. While these methods can identify surface-level problems to some extent, the information sources are often scattered. There is a lack of unified organization and correlation analysis between equipment data, environmental data, and personnel behavior information, leading to a fragmented understanding of the on-site safety status and hindering the formation of a holistic judgment consistent with the actual risk evolution process. This results in inaccurate risk identification and low safety levels in engineering projects. Therefore, improving the accuracy of risk identification and the safety of engineering projects has become an urgent technical problem to be solved. Summary of the Invention

[0003] The purpose of this invention is to design an engineering project safety management system that can improve the accuracy of risk identification and the safety of engineering projects.

[0004] To achieve the above objectives, a first aspect of the present invention provides an engineering project safety management system, the system comprising: The first calculation unit is used to acquire equipment status indicators, environmental change indicators, and personnel behavior indicators, and to perform weighted calculations based on the equipment status indicators, environmental change indicators, and personnel behavior indicators to obtain a comprehensive safety status result. The assessment unit is used to conduct risk assessment based on the comprehensive safety status result and the personnel behavior indicators, obtain a safety risk assessment result, and map the safety risk assessment result into a risk level identifier according to a preset multi-level boundary threshold. The second calculation unit is used to obtain the basic action intensity coefficient and reference upper boundary value corresponding to the risk level identifier, and to calculate the action intensity based on the basic action intensity coefficient, the reference upper boundary value and the safety risk assessment result to obtain the action intensity quantity. The management unit is used to generate a set of management instructions based on the risk level identifier, the action intensity, and the preset template hierarchy table, and to perform safety management and control on the project site based on the set of management instructions.

[0005] Further, the step of weighting the calculation based on the equipment status indicators, the environmental change indicators, and the personnel behavior indicators to obtain the comprehensive safety status result includes: Multiply the preset device weight by the device status index to obtain the first data; The second data is obtained by multiplying the preset environmental weights by the environmental change indicators; The third data is obtained by multiplying the preset behavioral weights by the personnel behavioral indicators; The first data, the second data, and the third data are added together to obtain the comprehensive security status result.

[0006] Furthermore, the step of conducting a risk assessment based on the comprehensive safety status results and the personnel behavior indicators to obtain a safety risk assessment result includes: Based on the comprehensive safety status results and the personnel behavior indicators, basic risk items and overall status and violation behavior superposition items are constructed respectively. The security risk assessment result is obtained by adding the basic risk items and the overall status with the violation items.

[0007] Furthermore, the construction of basic risk items and overall status and violation behavior superposition items based on the comprehensive security status results and personnel behavior indicators includes: The basic risk item is obtained by multiplying the preset behavioral impact coefficient by the personnel behavior index and adding the comprehensive safety status result. Multiply the comprehensive safety status result and the personnel behavior index to obtain the fourth data, and add the preset number to the personnel behavior index to obtain the fifth data; Divide the fourth data by the fifth data, and then multiply by a preset coupling enhancement coefficient to obtain the overall state and violation behavior superposition item.

[0008] Further, the step of calculating the motion intensity based on the basic motion intensity coefficient, the reference upper boundary value, and the safety risk assessment result to obtain the motion intensity quantity includes: A risk correction term is constructed based on the security risk assessment results and the preset continuous risk correction coefficients; A boundary approximation term is constructed based on the security risk assessment results, the reference upper boundary value, and the preset boundary approximation correction coefficient; The basic motion intensity coefficient, the risk correction term, and the boundary approximation term are added together to obtain the motion intensity quantity.

[0009] Furthermore, the step of constructing a risk correction term based on the security risk assessment results and preset continuous risk correction coefficients includes: Add the preset number to the security risk assessment result to obtain the sixth data. Divide the security risk assessment result by the sixth data and then multiply by the continuous risk correction coefficient to obtain the risk correction term.

[0010] Further, the step of constructing the boundary approximation term based on the security risk assessment result, the reference upper boundary value, and the preset boundary approximation correction coefficient includes: The boundary approximation term is obtained by dividing the safety risk assessment result by the reference upper boundary value and then multiplying it by the boundary approximation correction coefficient.

[0011] Furthermore, the management instruction set includes multiple management instructions. After performing safety management and control on the project site according to the management instruction set, the system further includes: The acquisition unit is used to acquire the execution record corresponding to each management instruction and generate the execution score corresponding to each execution record according to the preset execution score rules. The indicator unit is used to sum all the execution scores and divide by the total number of management instructions to obtain the execution consistency indicator.

[0012] Furthermore, after summing all the execution scores and dividing by the total number of management instructions to obtain the execution consistency index, the system further includes: The construction unit is used to construct the correction amount based on the action intensity, the execution consistency index, and the preset execution deviation amplification coefficient; An optimization unit is used to add the motion intensity amount and the correction amount to obtain an optimized motion intensity amount; wherein, the optimized motion intensity amount is used to generate management instructions for the next observation window.

[0013] Further, the step of constructing the correction amount based on the action intensity, the execution consistency index, and the preset execution deviation amplification coefficient includes: Subtract the preset number from the execution consistency index to obtain the seventh data; The correction amount is obtained by multiplying the seventh data, the action intensity, and the execution deviation amplification factor.

[0014] The beneficial technical effects of the present invention are at least as follows: To address the aforementioned issues, this invention provides an engineering project safety management system. Its core lies in constructing a safety management scheme based on the continuous control requirements in engineering project safety management, with unified expression of on-site status, coupled risk assessment, management instruction generation, and execution result correction as the main lines. This solution first addresses the problem of scattered sources and inconsistent expressions of information on abnormal equipment, environmental changes, and personnel violations at construction sites. It consolidates these three types of information into a unified, comprehensive safety status result, retaining personnel behavior indicators directly related to subsequent risk amplification. This ensures that subsequent judgments are based on a unified, compact, and directly calculable data foundation. Next, by combining the coupling relationship between the comprehensive safety status result and personnel behavior indicators, a quantitative assessment of safety risks within the current observation window is conducted. This ensures that the safety risk assessment results not only reflect the overall status level but also the amplifying effect of personnel violations in a high-risk context. After the safety risk assessment results are generated, corresponding action intensity quantities are generated based on the risk level identifier and the position of the safety risk assessment result within the current level range. Combined with action templates, resource mapping, and on-site control objects, this forms directly identifiable management instructions, transforming on-site control from simple alarm prompts into execution actions bound to specific personnel, equipment, and areas. After the management instructions are executed, execution consistency indicators are generated based on the execution records returned by on-site terminals, equipment interfaces, and area control units. These indicators are then used to adjust the action intensity for the next observation window, allowing subsequent control efforts to be dynamically adjusted according to on-site implementation. Through the above-mentioned technical concept, this invention realizes a continuous closed loop from safety information collection and risk identification to action execution and execution correction, so that the safety management of engineering projects no longer stops at static identification and manual response, but can continuously generate and correct control measures around the evolution of on-site risks, making it more suitable for full-process safety management applications in complex construction environments. Attached Figure Description

[0015] 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.

[0016] Figure 1 This is a schematic diagram of the structure of an engineering project safety management system provided in an embodiment of this application. Detailed Implementation

[0017] 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.

[0018] Please refer to Figure 1 , Figure 1 This is a schematic diagram of the structure of an engineering project safety management system provided in an embodiment of this application. The system includes: The first calculation unit 101 is used to acquire equipment status indicators, environmental change indicators and personnel behavior indicators, and to perform weighted calculations based on the equipment status indicators, environmental change indicators and personnel behavior indicators to obtain a comprehensive safety status result. Assessment unit 102 is used to conduct risk assessment based on comprehensive safety status results and personnel behavior indicators, obtain safety risk assessment results, and map the safety risk assessment results to risk level identifiers according to preset multi-level boundary thresholds. The second calculation unit 103 is used to obtain the basic action intensity coefficient and reference upper boundary value corresponding to the risk level identifier, and to calculate the action intensity based on the basic action intensity coefficient, reference upper boundary value and safety risk assessment results to obtain the action intensity quantity. Management unit 104 is used to generate a set of management instructions based on risk level identifiers, action intensity quantities, and preset template hierarchy tables, and to perform safety management and control on the engineering project site based on the set of management instructions.

[0019] In the first calculation unit 101, the safety management of the construction site does not involve indiscriminately collecting all site information. Instead, it prioritizes processing key information that can directly trigger management actions and transforms it into unified status data that can be directly accessed in subsequent steps. Equipment status, environmental changes, and personnel safety behavior are selected here because these three types of information correspond to mechanical operation risks, working environment risks, and personnel operation risks, respectively, covering the most significant sources of risk in project safety management.

[0020] During system deployment, raw equipment operation records are provided by status acquisition modules installed on tower cranes, construction hoists, suspended platforms, temporary power distribution devices, or concrete conveying equipment. The acquired data includes abnormal codes, status transition records, and fault trigger records output from existing control interfaces or additional acquisition nodes. For example, tower cranes can output hoisting mechanism abnormal codes, slewing resistance codes, and current fluctuation markers; construction hoists can output door lock abnormal signals, limit switch trigger records, and drive abnormal records; and temporary power distribution devices can output leakage alarms, overload alarms, and phase loss alarms. Raw environmental monitoring records are provided by environmental monitoring terminals deployed at the edge of the foundation pit, near scaffolding, material storage areas, around tower crane foundations, or at site access points. These terminals can use integrated acquisition boxes or aggregate and upload data using several independent sensor modules plus a local gateway. Recorded data enters the system in the form of a continuous sampling sequence. The original records of personnel safety behavior identification are jointly generated by on-site camera devices and edge analysis terminals. Camera devices are preferentially deployed at high-risk work entrances, material hoisting channels, scaffolding access points, near edges and openings, and large equipment operation areas. After receiving the video stream, the edge analysis terminal completes the identification locally and outputs event records, such as "entering the hoisting area without wearing a safety helmet", "entering the night work area without wearing a reflective vest", and "crossing the warning line to enter the restricted area". In this way, what enters the subsequent processing is no longer long video, but structured event data directly corresponding to the safety behavior.

[0021] After equipment status data enters the system, it is first uniformly organized according to the anomaly mapping table corresponding to the equipment type. Anomaly codes, status transitions, or alarm records uploaded by different devices are mapped to the same type of equipment anomaly event. Then, the occurrence of anomaly events is statistically analyzed according to a fixed observation window. Taking tower cranes as an example, hoisting overcurrent codes, slewing jam codes, and luffing anomaly codes are all mapped to equipment anomaly events; taking construction hoists as an example, door lock failure, landing door not closing, and overload triggering are also mapped to equipment anomaly events. The system counts the number of anomaly events within an observation window and corrects the results based on the concentration of anomaly occurrences, thereby obtaining the equipment status indicators. For example, if a tower crane experiences two hoisting anomalies and one slewing anomaly within a single observation window, the basic equipment anomaly value is recorded as 3. If these three anomalies occur within a very short period, the system identifies them as persistent anomalies, and the resulting value is adjusted accordingly. This is higher than typical discrete anomaly scenarios. After environmental data enters the system, the changes between adjacent sampling records are used as the processing object. The system compares continuous records and only accumulates the changes that reach the management concern threshold to form environmental change indicators. For example, if the wind speed at the edge of a foundation pit continuously increases across several sampling points, and the change crosses the designated monitoring range set by the project, then this increase will be accumulated into the project's data. When the humidity near a material storage yard increases continuously within a short period of time, and the increase reaches a preset threshold, the corresponding change is also included. After personnel safety behavior data enters the edge analysis terminal, the video stream is first extracted into continuous image frames, and then fed into a three-layer convolutional structure for recognition. The first layer extracts basic features such as head, protective clothing, and human body contours. The second layer extracts combined features such as "whether the safety helmet matches the head," "whether the reflective vest covers the upper body," and "whether the human body crosses the warning line." The third layer outputs the behavior category results. The system then uses a fixed observation window to count the number of violations, forming personnel behavior indicators. For example, if a camera is installed at the entrance to the area covered by the tower crane's slewing radius, and a worker enters the area twice through an observation window—once without a safety helmet and on another occasion crossing the temporary warning line into the hoisting waiting area—the edge terminal will output two violation records. The value is 2. After the above sorting, , and All of them are converted into structured indicators that can be directly used in subsequent calculations, among which This data was obtained from statistics on equipment malfunction events. Obtained from the accumulation of environmental changes, This data is derived from statistics on violations.

[0022] After obtaining the three types of indicators, the system performs uniform scaling on them according to the normalized intervals set during project initialization, and generates a comprehensive safety status result by combining the preset weights of the engineering scenario. The first data point is obtained by multiplying the preset equipment weights and equipment status indicators; the second data point is obtained by multiplying the preset environmental weights and environmental change indicators; the third data point is obtained by multiplying the preset behavioral weights and personnel behavioral indicators; and the third data point is then added together to obtain the comprehensive safety status result. The formula is shown below: ; in, The overall security status result is calculated by the system at the end of the current observation window; This is a device status indicator, and its value is derived from the statistical results of abnormal events of the device within this observation window; It is an environmental change indicator, and its value is derived from the cumulative change results between adjacent environmental records that have reached a set threshold. These are personnel behavior indicators, and their values ​​are derived from the statistical results of violations within this observation window. For device weights, For environmental weight, These are behavioral weights, pre-set at project initiation based on the project type. In an implementation scenario, if the normalized equipment status indicators... The value is 2, representing an environmental change index. The value is 1.5, representing a personnel behavior indicator. The value is 3, and at the same time , , If we take values ​​of 0.4, 0.3, and 0.3 respectively, the comprehensive safety status result calculated by the system will be... It is 2.15.

[0023] In different engineering projects, , , Different values ​​can be selected accordingly. For example, in projects where high-altitude hoisting operations account for a large proportion, and It can be set higher; in projects with a high proportion of deep foundation pits or open-air operations, This can be improved accordingly. After this processing, the system outputs a comprehensive safety status result. and personnel behavior indicators The former is used to characterize the overall safety status of the construction site within the current observation window, while the latter is used to separately identify personnel violations in subsequent risk identification, so that subsequent steps can directly generate judgments and management actions based on the structured results.

[0024] In evaluation unit 102, the comprehensive safety status result output by the above steps... The three categories of on-site factors—equipment malfunction, environmental changes, and personnel violations—have been compressed into an overall state quantity under the same calculation caliber. The above steps also retain personnel behavior indicators. This is used to separately characterize violations during the risk assessment phase. This step uses these two results as the sole input to further quantify the safety risks within the current observation window. The reason for this approach is that the actual risk at a construction site depends not only on the overall state but also on the context in which the violation occurs. Taking scenarios such as tower crane hoisting areas, foundation pit edges, and scaffolding access routes as examples, when equipment and the environment are already under pressure, the same instance of crossing the line, not wearing a safety helmet, or accidentally entering a hoisting area will have a higher probability of triggering an accident than under normal conditions. Therefore, the core of this step is to recombine the "overall state information" and "violation activity information" obtained from the first calculation unit into risk results that can directly drive management actions.

[0025] The basic form of risk calculation originates from the classical concept of risk superposition, which involves combining the contributions of independent risk sources using an additive structure. Following this approach, if only the direct effects of overall state and behavioral risks are considered, the basic risk can be written as... ,in This indicates the overall security status result. This indicates the direct increase in behavioral risk. This refers to the behavioral impact coefficient pre-set based on the project type at project initiation. While this basic form can express that "an increase in the number of violations increases risk," it is insufficient to capture the coupling mechanism commonly observed on construction sites where "the more tense the situation, the more dangerous the violation." Therefore, an interactive correction term is introduced on top of the basic risk. This interactive correction term comes from the common two-factor coupling modeling concept, meaning that the risk is only amplified when both the overall state and the violation exist simultaneously. Continue using it directly. While it can demonstrate a magnification effect, the growth is too rapid when there are many violations, which is not conducive to hierarchical management at the construction site. Therefore, drawing on the expression form of saturated growth, a denominator is added under the interactive correction term. The growth process is modified to a pattern of initial rapid increase followed by sustained but not excessively abrupt increases. Specifically, basic risk items and a combination of overall status and violation behavior items are constructed based on the comprehensive safety status results and personnel behavior indicators. The basic risk items are obtained by multiplying the preset behavioral impact coefficient by the personnel behavior indicators and adding the comprehensive safety status results. Multiply the comprehensive safety status result and personnel behavior indicators to obtain the fourth data point; add the preset number to the personnel behavior indicators to obtain the fifth data point; where the preset number is 1; divide the fourth data point and the fifth data point, and then multiply by the preset coupling enhancement coefficient to obtain the overall status and violation behavior superposition item. The safety risk assessment result is obtained by adding the basic risk items and overall status to the overlay items of violations, as shown in the following formula: ; in, This indicates the security risk assessment result for the current observation window; It represents the overall safety status result, and its value is derived from the fusion of equipment status indicators, environmental change indicators, and personnel behavior indicators. This represents a personnel behavior indicator, the values ​​of which are derived from the statistics of violations and the results of standardized processing within the current observation window. This represents the behavioral impact coefficient, which is pre-set in the project configuration library and used to control the direct impact of behavioral indicators on risk. This represents the coupling enhancement coefficient, also pre-configured by the project configuration library, used to control the amplification magnitude when the overall state and behavioral indicators work together. Because and On the same numerical scale and Taking dimensionless values, the parts of the above formula can be directly added together. The derivation of this formula is also relatively straightforward: the first term... Indicates overall background risk; the second item Indicates the direct increment of behavioral risk; the third item This represents the coupling increment when the overall state and the violation are superimposed, in the denominator. It is used to smooth the growth process of coupling terms, so that the model is closer to the change pattern of risk gradually increasing with the frequency of violations when used in the field.

[0026] When this formula is substituted into engineering scenarios for calculation, the logic of risk outcome changes will better align with on-site experience. Taking the hoisting operation area during the main structure construction phase as an example, if the overall safety status results... Personnel behavior indicators The behavioral impact coefficient of project configuration Coupling enhancement coefficient The current security risk assessment result of the observation window is: If, under the same overall condition, more violations are continuously identified, then... Upgrading to version 3.0 will change the security risk assessment results. Taking the perimeter area of ​​the foundation pit as another example, if the overall safety status has already reached a certain level due to environmental changes... Even if the personnel behavior indicators are only Under the same parameter configuration, it will also get From these calculation processes, it can be seen that the result obtained in this step... It can reflect the overall status that has been summarized, and it can also amplify high-risk conditions more sensitively when violations occur. This is especially important for safety management of engineering projects, because what really needs priority intervention on site is often the observation window where "the background status is already high and violations occur simultaneously".

[0027] After obtaining the safety risk assessment results Subsequently, the system continues to map the risk level identifier to the multi-level boundary thresholds in the project configuration library. This grading process uses segmented interval judgment, derived from the threshold grading method in engineering management. First, several risk boundaries are determined based on historical project cases, on-site trial operation records, and management requirements. Then, a fixed grade is assigned according to the interval the result falls into. Specifically, it is expressed as follows: ; in, Indicates the risk level identifier, based on The risk interval is obtained by looking up the table (risk interval table); This indicates the three-level boundary threshold preset by the project configuration library, which meets the requirements. These three thresholds are derived from the risk configuration during the project initiation phase and can be set in conjunction with the project type, work phase, historical hazard records, and management strategies. For example, in projects involving intensive high-altitude operations, The threshold can be lowered overall, allowing the system to reach a higher level earlier; during ordinary interior decoration stages, the threshold can be appropriately raised to match the lower overall risk background of the site. Continuing with the aforementioned embodiment, if the thresholds configured for a certain project are respectively... , , Then when At that time, the system determines Obtain risk level label ;when At that time, the system determines Obtain risk level label When the foundation pit is near the edge At that time, the system determines Obtain risk level label This shows that the same violation can be classified into different levels depending on the overall context, which corresponds to the management principle in engineering projects that "the same behavior has different levels of danger under different working conditions".

[0028] Following the above processing logic, this step ultimately outputs two results. The first is the security risk assessment result. The first is used to represent a continuous value indicating the risk intensity within the current observation window; the second is a risk level identifier. This is used to represent the discrete level of the current risk and provides a direct basis for generating specific management actions in the next step. In this way, the comprehensive safety status results and personnel behavior indicators obtained from the first calculation unit are further transformed into safety risk assessment results that retain calculation details and can be directly linked to management responses, realizing the transition from on-site data expression to on-site risk judgment.

[0029] In the second computing unit 103, the security risk assessment results and risk level label The risk status within the current observation window of the construction site has been expressed as a continuous result and a discrete hierarchical result. Subsequent steps use these two results as the sole input to further generate management decisions that can be directly executed on-site. These decisions are not simply fixed actions derived from a table based on risk levels, but rather, they refine the intensity and combination of management actions based on the position of the continuous risk result within the current risk level range. This ensures that even within the same risk level, the on-site management principle of "strengthening control in advance when approaching the boundary of the next higher level" is reflected. This approach is applicable to real-world scenarios in engineering project safety management, such as tower crane hoisting areas, deep foundation pit edges, scaffolding access points, and areas with concentrated temporary power supply. Under the same risk level, if the risk result is approaching the boundary of a higher level, on-site handling typically no longer requires ordinary inspections but should simultaneously involve dispatching personnel, restricting equipment, and controlling access to the area.

[0030] Specifically, the intensity of the movement Its construction originates from the hierarchical control concept and proportional correction concept in engineering control. The hierarchical control concept provides the basic form of "first determining the basic control intensity based on the level," therefore, the basic expression can be written first. ,in Risk level label The corresponding basic action intensity coefficient. The proportional correction concept is used to express that "the higher the continuous risk outcome, the stronger the control." If only linear correction is applied, it can form... ,in This represents the risk adjustment factor. However, at the engineering project site, the safety risk assessment results... In higher ranges, it is not advisable to cause the intensity of the movement to increase indefinitely; therefore, the linear correction should be adjusted accordingly. Rewritten in saturated growth form This ensures that while the intensity of actions continues to increase as the risk outcome rises, the rate of increase gradually stabilizes. Further consideration is given to risk level indicators. An interval division has already been given, and since states closer to the boundary of the next higher level within the same level should be subject to enhanced control in advance, a boundary approximation term is introduced. Among them, when , , hour, Take the upper boundary threshold of the corresponding risk interval respectively; when hour, The pre-set upper limit value for Level 4 risk control is taken from the project configuration library. This upper limit value is set during the project initiation phase based on the project type, historical high-risk records, and management tolerance range. It serves as a reference boundary to characterize the continued increase in action intensity under Level 4 risk conditions. Thus, the action intensity quantity gradually evolves from a basic form into an expression that simultaneously includes the level base value, continuous risk correction, and boundary approximation correction. Specifically, a risk correction term is constructed based on the safety risk assessment results and the pre-set continuous risk correction coefficient. The pre-set number is added to the safety risk assessment results to obtain the sixth data point. The safety risk assessment results are then divided by the sixth data point and multiplied by the continuous risk correction coefficient to obtain the risk correction term. A boundary approximation term is constructed based on the safety risk assessment results, the reference upper boundary value, and the pre-set boundary approximation correction coefficient. The safety risk assessment results are then divided by the reference upper boundary value and multiplied by the boundary approximation correction coefficient to obtain the boundary approximation term. The basic action intensity coefficient, the risk correction term, and the boundary approximation term are then added together to obtain the action intensity quantity, as shown in the following formula: ; in, This indicates the intensity of the action corresponding to the current observation window; This indicates the results of the safety risk assessment, and its value is derived from the risk calculation after coupling the comprehensive safety status results and personnel behavior indicators. Indicates the risk level; Indicates risk level label The corresponding basic action intensity coefficient is read from the level configuration table in the project configuration library. For example, when , , , At different times, different basic control levels can be applied; This represents the continuous risk correction factor, which is configured based on the project type during the project initiation phase. This represents the boundary approximation correction factor, used to increase the intensity of action when the risk outcome approaches the reference boundary of the current level; Risk level label The corresponding reference upper boundary value, when , , The time is read from the risk interval table, when The upper limit value of the level four risk control is read from the project configuration library.

[0031] The three parts in the above formula are logically stacked layer by layer. The first part... It specifies the minimum level of control required at the current risk level, ensuring that once a risk is assessed at a certain level, the site will at least enter a management state matching that level. Part Two Based on this, a smoothing correction for continuous risk is introduced, allowing different risk values ​​within the same level to correspond to different action intensities. Part Three This further expresses the boundary approximation effect, meaning that the closer the risk outcome is to the upper boundary of the current level, the closer the control actions are to the state before the upgrade. Taking the main structure construction hoisting area as an example, if... , The basic action intensity coefficient for level three risks in the project configuration Continuous risk correction coefficient Boundary approximation correction coefficient The upper boundary value is used as a reference for the level 3 risk range. Then there is If within the same level ,but Both are at level three risk, but the former is closer to the level four boundary, therefore requiring a higher level of physical exertion. If the risk level is raised to... At the same time, the intensity coefficient of the fourth level basic movement in the project configuration has been increased to Furthermore, since the upper limit of the Level 4 risk control is preset as the corresponding reference boundary in the project configuration library, a higher value can still be calculated under similar risk values. This leads to stronger on-site control actions. This calculation process shows that the action intensity inherits both the continuity of the safety risk assessment results and the hierarchical nature of the risk level identification, enabling more precise on-site decision-making.

[0032] In management unit 104, the motion intensity quantity is obtained. Then, the system generates a set of management instructions based on the action template table and the site resource mapping table in the project configuration library. This approach employs the rule-matching concept commonly found in engineering management, and adds a "strength-driven template overlay" mechanism. The basic form of rule-matching is "if a certain level is met, then the corresponding template group is triggered"; this step further specifies that within the same level, when... As the system enters different intensity ranges, more control actions are layered on top of the basic template. To this end, the system pre-sets several template levels for each risk level during the project initiation phase. For example, in a three-layer template system, the first layer includes the arrival of the responsible safety officer and on-site reminders; the second layer adds restrictions on associated equipment; and the third layer further adds risk area control. The system reads the current... After the corresponding risk template group, then according to The current template level should be selected based on the intensity range, and the personnel management system, equipment control interface, and area control system are invoked to complete the object mapping. Risk level identifier. Used to determine which template to enter, motion intensity. This is used to determine which layer to activate within a set of templates. The template level refers to the execution hierarchy within the same risk template group, arranged from low to high action intensity. It typically manifests as a progressively increasing combination of actions, such as basic reminders, enhanced restrictions, and coordinated control, used to differentiate the appropriate level of control measures to take at the same risk level. The selection of intensity range template levels refers to first pre-setting several intensity ranges for each risk template group, and then setting the current intensity range as the template level. The interval that falls into is matched with the corresponding level; if If it falls within a lower range, select the basic level. Entering a higher level involves selecting a higher level with more layered control actions. The action template table determines which actions should be executed at the current template level, while the site resource mapping table binds these actions to specific execution objects. Therefore, the system first... Determine the risk template group, and then based on After determining the template hierarchy, the corresponding action items for that hierarchy are retrieved from the action template table. These are then combined with the on-site resource mapping table to generate sets of personnel scheduling, equipment control, and area management instructions corresponding to the action items. To enable structured output of this process, a set of management instructions is generated. Express it in the following form: ; in, This represents the set of management instructions generated by the current observation window; Indicates the first Management instructions; Indicates the risk level label and intensity of movement The number of instructions to be generated under the combined effect is determined by the template hierarchy table in the project configuration library. Specifically, the template hierarchy table determines which action items should be enabled at the current level. The number of instructions depends on how many action items are configured at this level and how many specific execution instructions need to be generated after these action items are mapped. The template hierarchy table pre-specifies which action items are included at this level, such as whether it includes personnel presence, equipment speed limit, area control, broadcast reminder, etc. The system retrieves the action item list based on this; then, the field resource mapping table maps each action item to a specific execution object. For example, "personnel presence" is mapped to a safety officer, "equipment speed limit" is mapped to a tower crane controller, and "area control" is mapped to a gate or electronic fence. Finally, the preset action parameters and execution time limits are combined to assemble the content of each specific instruction. Each management instruction All are assembled by the system into a structure of "object number + action type + action parameters + execution time limit". For example, in a hoisting area scenario, one instruction could be "Safety officer AQ-07 arrives at area Z-13 within two minutes to conduct a confirmation inspection", another instruction could be "Tower crane controller TC-02 switches to the speed limit parameter group", and yet another instruction could be "Entrance gate G-13 enters restricted access state and links the warning screen to display a high-risk warning". Here, the object number comes from the site resource mapping table, the action type comes from the action template table, the action parameters come from the preset action parameter set in the project configuration library, and the execution time limit is determined by the risk level identifier and the action type. It is not an abstract "set of suggestions", but a set of field execution instructions that can be directly issued to mobile terminals, controllers or regional systems.

[0033] It should be noted that the above formula is a structured representation of the management instruction set. The key point is that the management instruction set consists of several management instructions, and the number of these instructions is influenced by both the risk level identifier and the action intensity. The role of the action template table and the on-site resource mapping table is reflected in the generation process of the management instruction set; that is, the system first... Determine the risk template group, and then based on First, determine the template hierarchy. Then, extract the corresponding action items from the action template table. Next, use the field resource mapping table to bind these action items to specific safety officers, equipment control interfaces, and area control objects, ultimately forming the set of management instructions represented in the formula. .

[0034] Specifically, the system first identifies the risk level. Determine which risk template group to enter, and then determine the intensity of the action. The system matches the intensity ranges within the template group to determine the appropriate template level to activate. The template level table pre-records the level numbers corresponding to different intensity ranges under each risk level. Based on this, the system obtains the current level and then reads the corresponding action item list from the action template table. Each action item list contains at least an action type identifier and an action object category identifier, such as "personnel arrival confirmation," "equipment speed limit," "area lockdown," and "broadcast notification." The action type indicates whether the instruction pertains to personnel dispatch, equipment control, or area management.

[0035] After obtaining the list of actions, the system calls the on-site resource mapping table to find the specific execution object based on the action object category identifier and the current risk area identifier, thereby determining the object number. If the action item is personnel arrival confirmation, the specific safety officer number is selected from the personnel resources bound to the current risk area, taking into account the duty status, responsibility area, and current location. If the action item is equipment speed limiting, the corresponding equipment control interface number is read from the equipment resources bound to the current risk area. If the action item is area lockdown or broadcast notification, the corresponding object number is read from the turnstile, electronic fence, broadcast terminal, or warning screen resources bound to the current risk area. In this way, the object number is determined through "action object category + risk area + on-site resource mapping relationship".

[0036] After determining the object number and action type, the system then reads the "risk level identifier" from the preset action parameter set. The action parameters correspond to the "+template level+action type". For example, personnel dispatch action parameters may include arrival confirmation method, patrol range, or confirmation requirements; equipment control action parameters may include speed limit gear, shutdown mode, or restricted range; area control action parameters may include lockdown level, broadcast content, or warning display content. Finally, the system identifies the risk level. The execution time limit is determined together with the action type, with higher risk levels and more critical action types corresponding to shorter execution time limits. At this point, each specific instruction can be assembled into a structured content of "object number, action type, action parameters, and execution time limit" and written into the management instruction set.

[0037] Continuing with the above-mentioned hoisting area scenario as an example, if the calculation has been obtained... The project configuration stipulates that when and It should be activated when it is in the high range. For the third-level template in the corresponding risk template group, the system will determine the three types of instructions to be generated based on the template hierarchy table: personnel dispatch instructions, equipment restriction instructions, and area control instructions. Subsequently, the system reads the location information and task occupancy status of the currently on-duty safety officer from the personnel management system, prioritizing the nearest safety officer who is not currently undertaking a higher-priority task, for example, generating the instruction "Safety officer AQ-07 arrives at area Z-13 within two minutes"; it reads the tower crane controller TC-02 bound to area Z-13 from the equipment control interface table and generates the instruction "TC-02 switches to speed limit mode"; it reads the entrance gate G-13 and broadcast terminal BC-13 from the area control system, correspondingly generating the instructions "G-13 restricts unauthorized personnel from entering" and "BC-13 continuously broadcasts risk warnings for the hoisting area". At this point, the management instruction set... It can contain three to four structured instructions, the exact number of which is given by the template hierarchy table. If the risk level is still [value] in another observation window... However, the intensity of the movement is only The system may only enable The second-level template in the corresponding risk template group generates two types of instructions: "safety officer on site + equipment restriction," without initiating area control actions. In this way, different risk locations within the same risk level can generate different combinations of management actions, making on-site handling more in line with the dynamic characteristics of engineering projects.

[0038] It should be noted that the first step is... Define risk template groups, for example Enter the Level 1 Risk Template Group. Enter the secondary risk template group. Enter the Level 3 risk template group. Enter the Level 4 Risk Template Group; the second step is... The template layer is compared with the preset intensity range within the template group to determine which layer to use. For example, this can be agreed upon. When the price falls into a low-level range, the base layer template, i.e., the first layer template, is activated. When the value falls within the median range, the reinforced template, i.e., the second template, is activated. When the price falls into the high-level range, the linked layer template, i.e., the third-level template, is activated. Thus, Decide which group to join. Decide "which layer to use".

[0039] After this process, the safety risk assessment results and risk level label This is further transformed into two results: one is the intensity of the movement. The first is used to quantify the level of control that should be executed within the current observation window; the second is a set of management instructions. This is used to concretely implement control measures down to the level of personnel, equipment, and specific areas. In this way, the entire solution achieves a shift from "risk assessment" to "management execution," allowing the field system to directly execute management instructions. Initiate actual control actions to achieve safety management and control of the engineering project site, and provide a unified input for the next step of on-site correction and continuous control based on the execution results.

[0040] In some embodiments, the action intensity quantity output by the aforementioned steps Management instruction set Having already concretized "how to handle current risks" into actionable on-site procedures, this step moves forward to address the questions of "whether these procedures are truly implemented, how well they are implemented, and how the control intensity should be adjusted in the next observation window." At the project site, while procedures such as safety officer arrival, equipment speed limiting, area lockdown, and broadcast reminders have been issued, the execution status of different types of procedures is not inherently consistent: some procedures take effect immediately, such as the equipment control interface returning a successful speed limit; some procedures have a delay, such as the safety officer needing to move to the designated area after receiving the instruction; and some procedures may only be partially completed, such as the gate entering a restricted state but the broadcast not yet being initiated. To ensure the entire solution forms a truly operational closed loop, this step consolidates the management instructions... Each execution record is converted into an execution record, and then these execution records are aggregated into a unified execution consistency indicator. Finally, based on this consistency index Action intensity After correction, the optimized motion intensity is obtained and can be directly used in the next observation window. In this way, the on-site status expression, safety risk assessment results, and management decisions will all be translated into quantifiable results of "whether the execution is effective" in this step, and the entire solution will thus be transformed from a risk identification system into a continuously adjustable safety management system.

[0041] Specifically, the set of management instructions Each management instruction in the system is generated with an object number, action type, action parameters, and execution time limit. Therefore, the acquisition unit first establishes an execution tracking table based on this information. The records in the execution tracking table correspond one-to-one with the instruction type. For personnel dispatch instructions, the system obtains the reception time, confirmation time, and arrival status through the safety officer's mobile terminal, and matches the location identifier reported by the terminal with the specified area in the instruction to obtain the execution record of that instruction. For equipment control instructions, the system returns whether the target parameters were successfully written, whether the current operating gear of the equipment has been switched to the target gear, and whether the target status maintenance time has met the requirements through the equipment control interface, forming the execution record of that equipment instruction. For area control instructions, the system confirms whether the area has entered a restricted state, whether the entrance is restricted as required, and whether the broadcast content has been started and is continuous through the electronic fence controller, the gate status interface, the broadcast terminal feedback, and the status words on the on-site warning screen, forming the execution record of that area instruction.

[0042] The system generates an execution score for each management instruction execution record according to the same execution scoring rule. The execution is scored as follows: 1 for complete compliance; 0 for completion within the specified observation window but with delays or local deviations, a preset score between 0 and 1 (e.g., 0.8 or 0.6); and 0 for non-execution or execution results significantly inconsistent with the target state. This approach is derived from the statistical concepts of weighted scoring and arithmetic mean, first mapping the execution results of heterogeneous objects to the same scoring interval, and then using the mean to reflect the overall execution status. The indicator unit is used to sum all execution scores and divide by the total number of management instructions to obtain the execution consistency index. If the management instruction set... The CCP If an instruction is executed, the execution consistency index of the current observation window will be [value]. The calculation method is as follows: ; in, This indicates the execution consistency metric for the current observation window; Represents a set of management instructions The total number of instructions in the set, this value is directly determined by the system. Counting yields the result; Indicates the first The execution score of each management instruction is derived from the score obtained by comparing the actual execution status of the instruction with the target execution status in the execution tracking table. The basic formula is the arithmetic mean formula, and its derivation logic is: first, the execution result of each instruction is uniformly expressed as a single score; then, the execution status of multiple instructions is compressed into an overall consistency index through the average. After this processing, The value remains within a uniform scale; the closer the value is to 1, the more fully the issued actions are implemented on-site; the closer the value is to 0, the worse the implementation. Taking a hoisting operation area as an example, if the management instruction set... The system contains four instructions: "Safety officer arrival confirmation," "Tower crane switches to speed limit gear," "Entrance gate enters restricted state," and "Broadcast terminal starts notification." After tracking this step, the execution scores for the four instructions are 1.0, 1.0, 0.6, and 0.8, respectively. Therefore, the execution consistency index for the current observation window is [insert value here]. For example, in a scenario involving the edge of a foundation pit, if two out of five instructions are fully executed, one is delayed, and two are not executed, then five execution scores can be generated, such as 1.0, 1.0, 0.5, 0, and 0. This indicates that the current level of implementation of control measures on-site is relatively low.

[0043] After obtaining the consistency index Then, the system continues to adjust the motion intensity based on the proportional feedback principle in control theory. Correction is performed. The basic form of proportional feedback can be expressed as "corrected control quantity = original control quantity + deviation term", where the deviation term is proportional to the "difference between the target state and the actual state". Mapped to this scheme, the target state can be understood as the management command being fully executed, corresponding to an execution consistency index close to 1; the actual state is calculated from the current observation window. Characterization, therefore, the execution bias can be written as The building block further constructs the deviation term into a value related to the motion intensity. The proportional correction amount, i.e. ,in To apply the deviation amplification factor, the optimization unit adds this correction amount to the original motion intensity amount to obtain the optimized motion intensity amount. After rearranging the addition form, it can be written as a multiplication structure that is easier for the system to implement: ; in, This represents the optimized motion intensity, used for generating management instructions in the next observation window; Indicates the intensity of the action; This indicates the execution consistency metric for the current observation window; This represents the execution deviation amplification factor, which is preset by the project configuration library at project startup based on the project type and management strategy. The derivation of this formula is very clear: it begins with the classic proportional feedback expression. Start by extracting common factors. ,get This structure makes it possible when When it approaches 1, Approaching 0, optimized motion intensity Compared with the original motion intensity Basically consistent; when At lower levels, Increase the intensity, and the system automatically amplifies the control strength of the next observation window. Continuing with the above example of the hoisting operation area, if the original motion intensity is... This step calculates the execution consistency index. The execution deviation amplification factor configured for the project is: Then there is This result indicates that in the next observation window, the system will generate new management instructions with a higher intensity of action than before. For example, it might increase the requirement from a single safety officer to two safety officers working together for confirmation, or it might overlay area control actions while maintaining equipment speed limits. In another scenario, if the original intensity of action remains the same... However, due to poor on-site execution conditions, the calculations yielded... In the same Below, there is This indicates that the next observation window will enter a more advanced control template. If all instructions are completed as required, Then there is This indicates that the current control level is matched with the on-site execution situation, and the system can directly maintain the original control level.

[0044] Through the above processing, this step ultimately yields two direct results: first, the consistency index is achieved. Its source is a set of management instructions. The results of each execution are scored and then averaged; secondly, the optimized motion intensity is... Its source is the original motion intensity. By implementing consistency indicators The result after proportional feedback correction. Thus, the management instruction set... The risks identified, quantified, and reinstated in this step are then applied to the next round of control processes, thus forming a continuous closed loop in project safety management: "risk assessment – ​​action generation – execution tracking – intensity correction." In field applications, this closed loop continuously mitigates the common problem of "instructions being issued but insufficient implementation on-site," ensuring the system doesn't merely remain at the alarm and suggestion level, but automatically increases or maintains control based on actual execution results, thereby more stably keeping risks within a manageable range.

[0045] The engineering project safety management system illustrated in this application obtains equipment status indicators, environmental change indicators, and personnel behavior indicators through a first calculation unit. It then performs a weighted calculation based on these indicators to obtain a comprehensive safety status result. An assessment unit performs a risk assessment based on the comprehensive safety status result and personnel behavior indicators, obtaining a safety risk assessment result. This result is then mapped to a risk level identifier based on preset multi-level boundary thresholds. A second calculation unit obtains the basic action intensity coefficient and reference upper boundary value corresponding to the risk level identifier. Based on these coefficients, the result, and the safety risk assessment result, it calculates the action intensity quantity. A management unit generates a set of management instructions based on the risk level identifier, the action intensity quantity, and a preset template hierarchy table. This set of instructions is used to manage and control the safety of the engineering project site. This improves the accuracy of risk identification and the safety of the engineering project.

[0046] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A safety management system for engineering projects, characterized in that, The system includes: The first calculation unit is used to acquire equipment status indicators, environmental change indicators, and personnel behavior indicators, and to perform weighted calculations based on the equipment status indicators, environmental change indicators, and personnel behavior indicators to obtain a comprehensive safety status result. The assessment unit is used to conduct risk assessment based on the comprehensive safety status result and the personnel behavior indicators, obtain a safety risk assessment result, and map the safety risk assessment result into a risk level identifier according to a preset multi-level boundary threshold. The second calculation unit is used to obtain the basic action intensity coefficient and reference upper boundary value corresponding to the risk level identifier, and to calculate the action intensity based on the basic action intensity coefficient, the reference upper boundary value and the safety risk assessment result to obtain the action intensity quantity. The management unit is used to generate a set of management instructions based on the risk level identifier, the action intensity, and the preset template hierarchy table, and to perform safety management and control on the engineering project site based on the set of management instructions.

2. The engineering project safety management system according to claim 1, characterized in that, The step of weighting and calculating the comprehensive safety status result based on the equipment status indicators, the environmental change indicators, and the personnel behavior indicators includes: Multiply the preset device weight by the device status index to obtain the first data; The second data is obtained by multiplying the preset environmental weights by the environmental change indicators; Multiply the preset behavioral weights by the personnel behavioral indicators to obtain the third data; The first data, the second data, and the third data are added together to obtain the comprehensive security status result.

3. The engineering project safety management system according to claim 1, characterized in that, The risk assessment based on the comprehensive safety status results and the personnel behavior indicators, to obtain the safety risk assessment results, includes: Based on the comprehensive safety status results and the personnel behavior indicators, basic risk items and overall status and violation behavior superposition items are constructed respectively. The security risk assessment result is obtained by adding the basic risk items and the overall status with the violation items.

4. The engineering project safety management system according to claim 3, characterized in that, The construction of basic risk items and overall status and violation behavior superposition items based on the comprehensive safety status results and personnel behavior indicators includes: The basic risk item is obtained by multiplying the preset behavioral impact coefficient by the personnel behavior index and adding the comprehensive safety status result. Multiply the comprehensive safety status result and the personnel behavior index to obtain the fourth data, and add the preset number to the personnel behavior index to obtain the fifth data; Divide the fourth data by the fifth data, and then multiply by a preset coupling enhancement coefficient to obtain the overall state and violation behavior superposition item.

5. The engineering project safety management system according to claim 1, characterized in that, The step of calculating the motion intensity based on the basic motion intensity coefficient, the reference upper boundary value, and the safety risk assessment result to obtain the motion intensity quantity includes: A risk correction term is constructed based on the security risk assessment results and the preset continuous risk correction coefficients; A boundary approximation term is constructed based on the security risk assessment results, the reference upper boundary value, and the preset boundary approximation correction coefficient. The basic motion intensity coefficient, the risk correction term, and the boundary approximation term are added together to obtain the motion intensity quantity.

6. The engineering project safety management system according to claim 5, characterized in that, The step of constructing a risk correction term based on the security risk assessment results and preset continuous risk correction coefficients includes: Add the preset number to the security risk assessment result to obtain the sixth data. Divide the security risk assessment result by the sixth data and then multiply by the continuous risk correction coefficient to obtain the risk correction term.

7. The engineering project safety management system according to claim 5, characterized in that, The step of constructing a boundary approximation term based on the security risk assessment result, the reference upper boundary value, and a preset boundary approximation correction coefficient includes: The boundary approximation term is obtained by dividing the safety risk assessment result by the reference upper boundary value and then multiplying it by the boundary approximation correction coefficient.

8. The engineering project safety management system according to claim 1, characterized in that, The management instruction set includes multiple management instructions. After performing safety management and control on the project site according to the management instruction set, the system further includes: The acquisition unit is used to acquire the execution record corresponding to each management instruction and generate the execution score corresponding to each execution record according to the preset execution score rules. The indicator unit is used to sum all the execution scores and divide by the total number of management instructions to obtain the execution consistency indicator.

9. The engineering project safety management system according to claim 8, characterized in that, After summing all the execution scores and dividing by the total number of management instructions to obtain the execution consistency index, the system further includes: The construction unit is used to construct the correction amount based on the action intensity, the execution consistency index, and the preset execution deviation amplification coefficient; An optimization unit is used to add the motion intensity amount and the correction amount to obtain an optimized motion intensity amount; wherein, the optimized motion intensity amount is used to generate management instructions for the next observation window.

10. The engineering project safety management system according to claim 9, characterized in that, The step of constructing the correction amount based on the action intensity, the execution consistency index, and the preset execution deviation amplification coefficient includes: Subtract the preset number from the execution consistency index to obtain the seventh data; The correction amount is obtained by multiplying the seventh data, the action intensity, and the execution deviation amplification factor.

Images