Construction site safety behavior identification method and system based on area driving

Abstract

The invention discloses a construction site safety behavior recognition method and system based on area driving, and relates to the technical field of data processing, and the method comprises the steps: obtaining a target construction site composed of recognition areas, acquiring an environment change data sequence, an equipment state data sequence and a personnel position data sequence in a previous acquisition time period of the current acquisition time period; obtaining an ith overrun index, and obtaining an ith environment change index; acquiring a danger coefficient of the identification area according to the ith environment change index and the ith overrun index; and obtaining a to-be-identified constructor, obtaining an ith position change trend value of the to-be-identified constructor, and obtaining a safety behavior identification result of the to-be-identified constructor according to the ith position change trend value of the to-be-identified constructor and the danger coefficient of the ith constructor in the identification area at the current moment. The method has the advantages of multi-risk fusion analysis, cross-regional collaborative early warning prediction and active risk-related behavior identification.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and specifically to a method and system for identifying safety behaviors at construction sites based on region-driven approaches. Background Technology

[0002] In complex construction site safety management, such as the safety monitoring of construction activities at large building construction sites, existing methods suffer from problems such as lagging dynamic risk perception and insufficient collaborative early warning capabilities.

[0003] Specifically, existing technologies mainly rely on static monitoring equipment (such as fixed cameras) or isolated sensors (such as independent pressure detectors), making it difficult to achieve real-time fusion analysis of multi-dimensional risk factors. Assessments of individual worker safety are typically limited to single-threshold alarms based on helmet wearing status (e.g., triggered by pressure exceeding limits), neglecting the potential escalation of risks implied by continuous external pressure trends (e.g., progressive compression). Secondly, the dynamic correlation between sudden environmental anomalies (e.g., sudden spatial distance changes caused by equipment tipping or material slippage) and personnel behavior is not effectively captured, leading to an inability to predict the impact of sudden hazards on surrounding areas. This is especially true in multi-tasking scenarios where the risk states of different areas influence each other. When a certain area becomes highly risky due to sudden environmental changes or abnormal personnel equipment, construction workers in neighboring areas may actively approach the dangerous area due to operational inertia. However, existing technologies cannot quantify the real-time diffusion coefficient of the hazard source, nor can they predict proactive risk-taking behavior (e.g., approaching the dangerous area) through personnel movement trajectories. Summary of the Invention

[0004] To address the technical problems of high false alarm / false alarm rates, inability to intervene in the trends of proactive human behavior, and inability to dynamically identify the combined risks of "high-risk areas - human approach behavior" in existing technologies, this invention provides a region-driven method and system for identifying safety behaviors at construction sites.

[0005] A region-driven method for identifying safety behaviors at construction sites includes: acquiring a target construction site composed of multiple consecutive identification regions; acquiring environmental change data sequences for each identification region within the previous acquisition time period, as well as equipment status data sequences and personnel location data sequences for each construction worker, based on infrared ranging sensors deployed in each identification region, pressure sensors inside the safety helmets of construction workers, and positioning wristbands deployed in each identification region; acquiring the i-th exceedance index based on the equipment status data sequence of the i-th construction worker, acquiring the identification region where the i-th construction worker is located at the current moment, and determining the environmental status of the identification region where the i-th construction worker is located at the current moment. The i-th environmental change index is obtained from the change data sequence; the hazard coefficient of the i-th construction worker in the identification area at the current moment is obtained based on the i-th environmental change index and the i-th exceedance index; construction workers who are not in the same identification area as the i-th construction worker at the current moment are identified as construction workers to be identified, and the i-th position change trend value of the construction workers to be identified is obtained based on the personnel position data sequence of the construction workers to be identified and the identification area at the current moment of the i-th construction worker; and the safety behavior identification result of the construction workers to be identified is obtained based on the i-th position change trend value of the construction workers to be identified and the hazard coefficient of the identification area at the current moment of the i-th construction worker.

[0006] Optionally, obtaining the i-th exceedance index based on the equipment status data sequence of the i-th construction worker includes: if there is a pressure value exceeding the pressure threshold in the equipment status data sequence of the i-th construction worker, then the difference between the maximum pressure value and the pressure threshold is obtained as the first exceedance value; the change between the next pressure value and the previous pressure value in the equipment status data sequence of the i-th construction worker is obtained, and all changes are added together to obtain the second exceedance value; if the second exceedance value is less than zero, the first exceedance value is added to zero to obtain the i-th exceedance index; if the second exceedance value is not less than zero, the first exceedance value and the second exceedance value are added together to obtain the i-th exceedance index.

[0007] Optionally, obtaining the i-th environmental change index based on the environmental change data sequence of the identification area where the i-th construction worker is located at the current moment includes: determining whether there is fluctuation in the environmental change data sequence of the identification area where the i-th construction worker is located at the current moment, wherein the fluctuation is defined as the absolute change amplitude between adjacent distance data in the sequence exceeding a preset change threshold; if there is fluctuation, then a first preset value is obtained and used as the i-th environmental change index; if there is no fluctuation, then a second preset value is obtained and used as the i-th environmental change index.

[0008] Optionally, obtaining the risk coefficient of the identification area where the i-th construction worker is located at the current moment based on the i-th environmental change index and the i-th exceedance index includes: using the i-th exceedance index as the basic coefficient, using the product of the i-th exceedance index and the i-th environmental change index as the additional coefficient, and using the sum of the basic coefficient and the additional coefficient as the risk coefficient of the identification area where the i-th construction worker is located at the current moment.

[0009] Optionally, obtaining the i-th position change trend value of the construction worker to be identified based on the personnel location data sequence of the construction worker to be identified and the identification area where the i-th construction worker is located at the current moment includes: obtaining the relative distance value between each location data in the personnel location data sequence of the construction worker to be identified and the identification area where the i-th construction worker is located at the current moment, and forming a relative distance sequence of the construction worker to be identified according to the order of the location data in the location data sequence; and taking the difference between the next relative distance value and the previous relative distance value in the relative distance sequence of the construction worker to be identified, and taking the sum of all differences in the relative distance sequence as the i-th position change trend value of the construction worker to be identified.

[0010] Optionally, obtaining the safety behavior identification result of the construction worker to be identified based on the trend value of the i-th position change of the construction worker to be identified and the danger coefficient of the identification area where the i-th construction worker is located at the current moment includes: if the trend value of the i-th position change of the construction worker to be identified is less than zero, then the product of the trend value of the i-th position change of the construction worker to be identified and the danger coefficient of the identification area where the i-th construction worker is located at the current moment is used as the target identification value, and the behavior identification result of the construction worker to be identified is obtained based on the target identification value; if the trend value of the i-th position change of the construction worker to be identified is greater than zero, then the behavior identification result of the construction worker to be identified as a safe behavior is obtained.

[0011] A region-driven construction site safety behavior recognition system is also provided. The system includes: an acquisition module, used to acquire a target construction site composed of multiple continuous recognition areas, and based on infrared ranging sensors deployed in each recognition area, pressure sensors inside the safety helmets of construction workers, and positioning wristbands of construction workers, to acquire environmental change data sequences of each recognition area within the previous acquisition time period, as well as equipment status data sequences and personnel location data sequences of each construction worker; and a personnel data processing module, used to acquire the i-th exceedance index based on the equipment status data sequence of the i-th construction worker, and to acquire the recognition area where the i-th construction worker is located at the current moment, and to process the data based on the i-th construction worker's location within the recognition area at the current moment. The system obtains the i-th environmental change index from the environmental change data sequence; the regional data processing module obtains the hazard coefficient of the i-th construction worker's current location within the identification area based on the i-th environmental change index and the i-th exceedance index; the behavior recognition module identifies construction workers not located in the same identification area as the i-th construction worker and designates them as the construction workers to be identified, obtains the i-th location change trend value of the construction workers to be identified based on their personnel location data sequence and the identification area where the i-th construction worker is located at the current moment, and obtains the safety behavior recognition result of the construction workers to be identified based on the i-th location change trend value and the hazard coefficient of the identification area where the i-th construction worker is located at the current moment.

[0012] Optionally, the personnel data processing module is further configured to: if there is a pressure value exceeding the pressure threshold in the equipment status data sequence of the i-th construction worker, then obtain the difference between the maximum pressure value and the pressure threshold and use it as the first over-limit value; obtain the change between the next pressure value and the previous pressure value in the equipment status data sequence of the i-th construction worker, and add all the changes in sequence to obtain the second over-limit value; if the second over-limit value is less than zero, then add the first over-limit value and zero to obtain the i-th over-limit index; if the second over-limit value is not less than zero, then add the first over-limit value and the second over-limit value to obtain the i-th over-limit index.

[0013] Optionally, the personnel data processing module is further configured to: determine whether there is fluctuation in the environmental change data sequence of the identification area where the i-th construction worker is located at the current moment, wherein the fluctuation is defined as the absolute change amplitude between adjacent distance data in the sequence exceeding a preset change threshold; if there is fluctuation, obtain a first preset value and use it as the i-th environmental change indicator; if there is no fluctuation, obtain a second preset value and use it as the i-th environmental change indicator.

[0014] Optionally, the regional data processing module is also used to: take the i-th exceeding limit indicator as the basic coefficient, take the product of the i-th exceeding limit indicator and the i-th environmental change indicator as the additional coefficient, and take the sum of the basic coefficient and the additional coefficient as the danger coefficient of the identification area where the i-th construction worker is located at the current moment.

[0015] The beneficial effects of this invention are reflected in: The entire region-driven construction site safety behavior identification method first integrates data from infrared ranging sensors, safety helmet pressure sensors, and positioning wristbands. It fuses and analyzes discrete environmental abrupt change signals (such as equipment displacement), personnel and equipment pressure trends (such as gradual compression), and real-time location information. This captures sudden environmental anomalies (such as sudden distance changes caused by material slippage) and identifies persistent individual risks (such as continuously increasing safety helmet pressure even if not exceeding limits). The method quantifies the regional risk level through dynamic calculation of the hazard coefficient (such as automatically increasing the hazard coefficient when environmental abrupt changes are combined with equipment pressure), significantly reducing the false negative rate caused by single threshold alarms or isolated risk analysis. Furthermore, it calculates the cross-regional movement trend of construction personnel based on historical location sequences. When personnel from other regions continuously approach high-risk areas (the position change trend value is negative), it generates a combined risk judgment value by combining the real-time hazard coefficient of the target area (such as a high-risk gas leak area + maintenance personnel continuously approaching), accurately identifying "proactive risk-taking behavior" and realizing the transformation from passive response to proactive intervention. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts are not necessarily drawn to scale.

[0017] Figure 1 This is a partial flowchart of the region-driven construction site safety behavior recognition method of the present invention; Figure 2 This is a schematic diagram of another part of the process of the region-driven construction site safety behavior recognition method of the present invention; Figure 3 This is a schematic diagram illustrating the steps of the region-driven construction site safety behavior recognition method of the present invention; Figure 4 This is a schematic diagram of part of step S2 in the region-driven construction site safety behavior recognition method of the present invention; Figure 5 This is a schematic diagram of another part of step S2 in the region-driven construction site safety behavior recognition method of the present invention; Figure 6This is a schematic diagram of part of step S4 in the region-driven construction site safety behavior recognition method of the present invention; Figure 7 This is a schematic diagram of another part of step S4 in the region-driven construction site safety behavior recognition method of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0019] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0020] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, the terms "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0021] like Figure 1 , Figure 2 and Figure 3 As shown, a region-driven method for identifying safety behaviors at construction sites is provided. In one embodiment, the method includes: S1. Acquire the target construction site consisting of multiple continuous identification areas, and based on the infrared ranging sensors deployed in each identification area, the pressure sensors in the safety helmets of construction workers, and the positioning wristbands of construction workers, acquire the environmental change data sequence of each identification area in the previous collection time period, as well as the equipment status data sequence and personnel location data sequence of each construction worker. S2. Obtain the i-th out-of-limit index based on the equipment status data sequence of the i-th construction worker, obtain the identification area where the i-th construction worker is located at the current moment, and obtain the i-th environmental change index based on the environmental change data sequence of the identification area where the i-th construction worker is located at the current moment. S3. Obtain the danger coefficient of the identification area where the i-th construction worker is located at the current moment based on the i-th environmental change index and the i-th exceedance index; S4. At the current moment, acquire construction workers who are not in the same identification area as the i-th construction worker and designate them as construction workers to be identified. Based on the personnel location data sequence of the construction workers to be identified and the identification area where the i-th construction worker is located at the current moment, obtain the i-th position change trend value of the construction workers to be identified. Based on the i-th position change trend value of the construction workers to be identified and the danger coefficient of the identification area where the i-th construction worker is located at the current moment, obtain the safety behavior identification result of the construction workers to be identified.

[0022] In this embodiment, it should be noted that S1 provides the basic input for subsequent dynamic risk analysis. Specifically, the target construction site is divided into multiple continuous geographical sub-regions. These sub-regions are intelligently divided according to the construction progress and the distribution of work types. For example, the high-rise building structural frame area, material storage area, and equipment operation area are set as interconnected identification areas to ensure coverage of all critical work points. Dedicated sensor equipment is deployed in each area: infrared ranging sensors are installed on fixed brackets or mobile devices to monitor changes in environmental distance in real time; safety helmets worn by construction workers have built-in pressure sensors to detect external forces; and positioning wristbands are attached to the personnel's wrists to continuously track their position. This multi-dimensional sensing deployment ensures the comprehensiveness of the data source, creating conditions for identifying key risk factors such as personnel and equipment status, environmental anomalies, and personnel location.

[0023] Furthermore, the data sequences acquired by S1 are all based on historical time periods, generating environmental change data sequences (comprising distance data), equipment status data sequences for each construction worker (comprising pressure data), and personnel location data sequences (comprising location data). Specifically, this includes infrared ranging sensor data sequences (reflecting changes in the regional environment), pressure sensor data sequences (reflecting the pressure state of safety helmets), and positioning wristband data sequences (reflecting personnel movement trajectories) continuously recorded in the time period preceding the current acquisition time. For example, the infrared ranging sequence captures distance fluctuations of equipment or materials within the area; the pressure sequence records continuous changes in the force exerted on the safety helmet; and the location sequence displays personnel displacement through coordinate changes. This design utilizes historical sequence data to support trend analysis (such as calculating pressure change trends or location convergence trends in subsequent steps), avoiding the delay issues caused by processing real-time burst data, thereby improving response efficiency and practicality, and providing a stable data foundation for subsequent real-time risk assessment.

[0024] In S2, the "i" in the i-th construction worker and subsequent occurrences of "i" are positive integers, less than the number of construction workers. The i-th construction worker is any single worker. Dynamic assessment of the individual risk status of construction workers and the anomalies in their environment provides input for subsequent regional risk aggregation. First, for the i-th construction worker, the data sequence formed by their helmet pressure sensor in the previous time period is analyzed: on the one hand, it detects whether there is an instantaneous pressure exceeding the limit (such as a heavy object hitting the helmet), recording the maximum exceeding value to reflect the severity of the impact; on the other hand, it calculates the cumulative change of adjacent pressure values ​​within the sequence. If a continuous upward trend is shown (such as gradually intensifying compression), it is considered a potential risk signal even if the current limit is not exceeded. These two types of indicators are superimposed to form an "over-limit indicator"—which includes both the intensity of sudden pressure and captures the trend changes of gradual risk. At the same time, the actual location of the person at the current moment (e.g., the tower crane operation area) is obtained through the positioning wristband, and associated with the infrared ranging data sequence of the area: by judging whether the sudden change in adjacent distance values ​​in the sequence reaches the preset threshold (e.g., the sudden reduction in distance caused by equipment tilting), a binary "environmental change index" is generated (1 if the sudden change exists, otherwise 0).

[0025] Furthermore, for example, if a worker is operating in a steel structure installation area and their safety helmet pressure sequence shows that the pressure value is not exceeding the threshold but continues to rise (indicating a gradual increase in load due to loose support), then the second type of change contributes significantly to the positive over-limit indicator. Meanwhile, the infrared ranging sequence detects a sudden change in distance caused by the displacement of a nearby steel beam, and the environmental indicator is set to 1. This design, by integrating the equipment pressure trend with real-time environmental change signals, overcomes the limitations of existing single-threshold alarms: if only the pressure threshold is relied upon, this scenario may be missed; if only environmental changes are analyzed, it is impossible to correlate it with the specific pressure state of the personnel. Accurate correlation of the current location (rather than the location of the previous time period) ensures the real-time nature of regional risk assessment and avoids misjudgments caused by the movement of construction personnel across areas.

[0026] In S3, by dynamically integrating individual personnel risks with sudden environmental anomalies, the real-time hazard coefficient of the identified area where construction workers are currently located is quantified. This step collaboratively calculates two types of key indicators obtained in S2—"over-limit indicators" reflecting the pressure trend on personnel and equipment (such as continuously escalating helmet compression) and "environmental change indicators" marking sudden environmental changes (such as sudden distance changes caused by equipment tipping over). The specific logic is as follows: the over-limit indicator is used as the base coefficient to reflect the direct risk level faced by the personnel themselves; then, the product of the over-limit indicator and the environmental change indicator is used as an additional coefficient to amplify the combined hazard when sudden environmental changes and personnel risks coexist. For example, in the concrete pouring area, if vibration causes formwork displacement (environmental indicator is 1) and the pressure on the worker's helmet continues to rise (significant over-limit indicator), the additional coefficient will increase sharply, warning of the superimposed risk of "formwork collapse + personnel compression".

[0027] Furthermore, S3 effectively addresses the problem of existing methods neglecting risk coupling. When the environmental indicator is 0 (no sudden anomalies), the hazard coefficient depends only on the exceedance indicator, reflecting the independent impact of progressive risks. However, when the environmental indicator is 1, the additional coefficient significantly increases the hazard level, highlighting the multiplier effect of sudden environmental events on personnel risks (such as material slippage escalating the risk to already stressed personnel). For example, if a welding spark suddenly ignites materials in a rebar processing area (a sudden environmental change), even if a welder's own helmet pressure is not exceeded (the exceedance indicator is low), the additional coefficient will still be activated due to the sudden environmental change, quickly identifying secondary risks triggered by external events, thus avoiding the omission of combined high-risk scenarios due to isolated analysis.

[0028] In S4, the goal is to achieve cross-regional collaborative risk early warning by dynamically capturing the behavioral trends of construction workers in other areas actively approaching high-risk areas. First, for the personnel to be identified (e.g., workers in a nearby rebar processing area) located in different identification areas from the i-th construction worker (e.g., a tower crane operator), their location sequence data from the previous time period is extracted, and the relative distance sequence between their location and the danger zone of the i-th personnel is calculated. A "location change trend value" is generated by summing the continuous distance differences in the sequence: if the value is negative (e.g., -5), it indicates that the worker is continuously shortening the distance to the danger zone, demonstrating obvious active approach behavior; if the value is positive, it indicates moving away, and no warning is needed.

[0029] Furthermore, this trend value is correlated with the coefficient of the current hazardous area (generated by S3): when the trend value is negative, its product with the hazard coefficient is used as the final risk assessment value. For example, if a gas leak occurs suddenly in a pipeline installation area (high hazard coefficient), and maintenance personnel in adjacent areas continue to approach the area due to work inertia (negative trend value), the product result will decrease sharply, amplifying the combined risk of "high-risk area + active approach"; conversely, if the hazard coefficient is low (e.g., only a minor tool fall), even if someone approaches, the assessment value is still within a controllable range. This design breaks through the limitations of existing static monitoring—it quantifies the real-time diffusion intensity of the hazard source (through the hazard coefficient) and predicts the intentions of personnel behavior through historical trajectories (through the trend value), thereby achieving precise interception of risky behaviors such as "workers still carrying materials to the collapse area when scaffolding collapses".

[0030] In summary, the entire region-driven construction site safety behavior identification method first integrates data from infrared ranging sensors, safety helmet pressure sensors, and positioning wristbands. It fuses and analyzes discrete environmental abrupt changes (such as equipment displacement), personnel and equipment pressure trends (such as progressive compression), and real-time location information. This captures sudden environmental anomalies (such as sudden distance changes caused by material slippage) and identifies persistent individual risks (such as continuously increasing safety helmet pressure even if not exceeding limits). The method quantifies regional risk levels through dynamic calculation of hazard coefficients (e.g., automatically increasing hazard coefficients when environmental abrupt changes are combined with equipment pressure), significantly reducing the false negative rate caused by single threshold alarms or isolated risk analysis. Furthermore, it calculates the cross-regional movement trend of construction personnel based on historical location sequences. When personnel from other regions continuously approach high-risk areas (negative position change trend value), it combines the real-time hazard coefficient of the target area to generate a combined risk judgment value (e.g., high-risk gas leak area + continuous approach of maintenance personnel), accurately identifying "proactive risk-taking behavior" and achieving a shift from passive response to proactive intervention. In summary, this approach not only solves the challenge of real-time collaborative analysis of multiple risk factors (such as avoiding assessment bias caused by construction workers moving across areas), but also effectively intercepts cascading risk scenarios that existing methods cannot predict (such as personnel moving towards the collapse area) by quantifying the correlation between the intensity of hazard spread and personnel behavioral intentions, providing global safety decision support for multi-trade cross-operations.

[0031] like Figure 4 As shown, in one embodiment, obtaining the i-th exceedance index based on the equipment status data sequence of the i-th construction worker in S2 includes: S21. If there is a pressure value exceeding the pressure threshold in the equipment status data sequence of the i-th construction worker, then the difference between the maximum pressure value and the pressure threshold is obtained and used as the first over-limit value. S22. Obtain the change between the next pressure value and the previous pressure value in the equipment status data sequence of the i-th construction worker, add all the changes in sequence and obtain the second over-limit value. S23. If the second over-limit value is less than zero, the first over-limit value and zero are added together to obtain the i-th over-limit index. If the second over-limit value is not less than zero, the first over-limit value and the second over-limit value are added together to obtain the i-th over-limit index.

[0032] In this embodiment, it should be noted that in S21, sudden equipment risks are captured. When analyzing the equipment pressure data sequence of the i-th construction worker, the system first checks whether there is an instantaneous pressure peak exceeding a preset safety threshold. If such an event exists (e.g., a heavy object falling and hitting a safety helmet), the maximum pressure value in the sequence is extracted, and its positive deviation from the safety threshold is calculated as the first over-limit value. This step is designed to address the shortcomings of existing single-threshold alarms: for example, when a steel beam hoisted by a tower crane accidentally detaches and hits a worker's safety helmet below, although the pressure value drops after the impact, the maximum impact value has already far exceeded the threshold. If only the real-time threshold is relied upon for judgment (the alarm is lifted when the pressure returns to normal after the impact), the actual severity of the injury cannot be recorded. The first over-limit value, by recording the maximum deviation, quantifies the energy intensity of the sudden event, providing basic parameters for subsequent superimposed risk analysis.

[0033] It should also be noted that the pressure threshold is first determined based on standards and specifications, referring to the impact absorption performance requirements in the national standard "Head Protection Helmets" (e.g., the force transmitted to the head should not exceed 4.9kN), combined with the material compressive strength data provided by the helmet manufacturer, to set a static pressure alarm threshold (e.g., threshold A). Then, dynamic adaptation is performed, based on the risk level of different jobs (e.g., the threshold for high-altitude work is 20% lower than that for ground work), and reverse calibration is performed using historical accident data (e.g., the pressure peak during crush injuries). The construction party must complete pressure calibration testing (e.g., simulating impacts from falling objects of 5kg to 30kg) before system deployment to ensure that the threshold covers the actual risk range.

[0034] In S22, the focus is on the gradual trend of equipment pressure. The difference in pressure values ​​at adjacent time points within the same pressure sequence is extracted, and the algebraic sum of all consecutive differences is used as the second over-limit value. This value reflects the direction and cumulative intensity of continuous pressure change: if the pressure at a later time point in the sequence is consistently greater than that at the previous time point (positive difference), the cumulative result will increase significantly; conversely, if the pressure fluctuates and decreases (negative difference), the cumulative result will decrease. For example, in tunnel construction, if the rock strata above a worker's head slowly shift, the pressure on the safety helmet continuously increases but does not reach the threshold (no alarm from single-point detection). In this case, the second over-limit value, by accumulating and increasing the difference, captures the potential trend of "continuous rock strata pressure." This mechanism overcomes the blind spot of existing static thresholds for gradual risks, and is particularly suitable for slow-progressing risk scenarios such as roof settlement and support deformation.

[0035] In S23, if the second exceedance value is negative (pressure fluctuations generally show a downward trend), it indicates that the risk does not show signs of continuous escalation. In this case, only the first exceedance value (sudden impact intensity) is retained as the final exceedance indicator. If the second exceedance value is non-negative (pressure is stable or continuously rising), the first and second exceedance values ​​are superimposed, encompassing both sudden intensity and trend increment. For example, during concrete pouring, a worker's safety helmet experiences two impacts (the first exceedance value is high), but the subsequent pressure value is stable. In this case, because the second exceedance value is zero, the final indicator only reflects the instantaneous impact. Conversely, if the pressure continues to rise after the impact (e.g., the deformation of the pouring formwork intensifies the load), the superimposed result significantly amplifies the risk level. This logic avoids misjudgments caused by isolated analysis: a high second exceedance value alone may be noise fluctuation, requiring the first exceedance value to verify the existence of actual risk.

[0036] like Figure 5 As shown, in one embodiment, obtaining the i-th environmental change index in S2 based on the environmental change data sequence of the identified area where the i-th construction worker is located at the current moment includes: S24. Determine whether there is fluctuation in the environmental change data sequence of the identification area where the i-th construction worker is located at the current moment, wherein the fluctuation is defined as the absolute change amplitude between adjacent distance data in the sequence exceeding a preset change threshold. S25. If there is fluctuation, obtain the first preset value and use it as the i-th environmental change indicator. S26. If there is no fluctuation, obtain the second preset value and use it as the i-th environmental change indicator.

[0037] In this embodiment, it should be noted that in S24, environmental anomalies are determined by detecting abrupt changes in adjacent distances. The absolute difference between any two consecutive data points in the infrared ranging sequence is analyzed. When any difference exceeds a preset stability threshold (such as a sudden change in distance caused by equipment tipping over), a "fluctuation" is determined to exist. This threshold filters out subtle changes (such as slight displacement of building materials caused by wind), capturing only significant changes in spatial structure. For example, in a large equipment installation area, if the infrared sequence shows that the distance between adjacent measurement points suddenly drops from 2 meters to 0.5 meters (below the preset 1-meter fluctuation threshold), it is determined to be a reliable signal of equipment tilting or material collapse. This mechanism avoids misjudging non-risk behaviors such as personnel movement or the movement of small tools as environmental anomalies, ensuring the clarity of the physical meaning of environmental abrupt change signals.

[0038] It should also be noted that the preset fluctuation threshold is first determined based on the nominal error of the infrared sensor model (e.g., ±2% of the range), setting a minimum effective fluctuation threshold (e.g., threshold B) to filter out equipment-specific errors (e.g., ±5cm fluctuations caused by temperature drift). Then, it is dynamically adjusted: in densely populated equipment areas (e.g., within the tower crane's operating radius), a low threshold is set (e.g., an alarm is triggered if the distance suddenly increases by >50cm) to quickly detect abnormal boom displacement; in material stacking areas, a higher threshold (e.g., >1m) is used to avoid false alarms due to natural settlement of the stack. The adjustment method involves continuously collecting 72 hours of normal operating data, calculating the root mean square error of the distance change, and setting the fluctuation threshold to 3 times the root mean square error (covering 99.7% of normal fluctuations).

[0039] In S25, detected fluctuations are assigned a binary risk label. When S24 confirms the existence of a valid fluctuation (such as a jump in ranging value caused by scaffold displacement), the environmental change index is set to 1, representing a sudden physical environmental anomaly in the area. For example, in a scenario of stack slippage in a storage area, the infrared ranging sequence captures a continuous and abnormal reduction in the distance between the stack and the fixed support, triggering the fluctuation label. This index is strongly correlated with the deployment characteristics of environmental equipment: the infrared sensor coverage area is fixed, and a value of 1 only indicates that an abnormal displacement requiring intervention has occurred in the area, without involving the specific degree of damage (the degree is quantified by the subsequent risk factor), but it provides key environmental context for the risk to associated personnel.

[0040] In S26, scenarios with no significant fluctuations are handled. If the difference between adjacent values ​​in the infrared ranging sequence does not exceed a preset threshold (such as minor distance fluctuations caused by equipment vibration), the environmental change index is set to 0, indicating that there is no sudden structural risk in the current area. For example, the infrared data in the tower crane operation area may fluctuate regularly due to the normal rotation of the tower arm, but because the amplitude of the change is lower than the threshold, the environmental index remains at 0. This setting avoids false alarms caused by sensor noise or interference from normal operations. At the same time, a value of 0 does not exclude the existence of slow deformation in the area (such as foundation settlement). Such risks rely on other means of monitoring, but provide a "no sudden change" baseline state for the environmental dimension of subsequent risk factor calculation.

[0041] In one implementation, S3, obtaining the danger coefficient of the identified area where the i-th construction worker is located at the current moment based on the i-th environmental change index and the i-th exceedance index, includes: The i-th exceedance index is used as the basic coefficient, and the product of the i-th exceedance index and the i-th environmental change index is used as the additional coefficient. The sum of the basic coefficient and the additional coefficient is used as the danger coefficient of the i-th construction worker in the identified area at the current moment.

[0042] In this implementation, it should be noted that the precise quantification of regional hazard levels is achieved by dynamically coupling individual personnel risk with sudden environmental changes. The calculation logic includes a dual enhancement mechanism: First, an exceedance index is used as the base coefficient. This index is generated in stage S2 by integrating sudden pressure exceedances of construction personnel and equipment (such as heavy object impacts) and sustained pressure trends (such as slow roof settlement). For example, in high-altitude welding operations, if the pressure sequence of a worker's safety helmet shows a continuous increase even though it does not exceed the threshold (reflecting the risk of fatigue deformation of the steel beam), the exceedance index will capture this progressive hazard by accumulating and increasing the difference, thus separately characterizing the individual personnel risk.

[0043] Secondly, an additional coefficient is introduced by multiplying the environmental change index (binary variable) and the exceedance index. The core of this design lies in identifying the superimposed effect of "coexistence of environmental mutation and personnel risk": when the environmental index is 1 (such as a sudden drop in infrared ranging in the tower crane area, indicating abnormal boom sway), the additional coefficient is automatically activated and exponentially amplifies the basic risk. For example, when a concrete pump truck hose bursts (environmental mutation), if nearby workers simultaneously experience increased pressure on their safety helmets due to material burial (exceeding the exceedance index), the additional coefficient will drastically increase the danger factor, warning of the compound disaster chain of "equipment failure + personal injury"; conversely, if there is no environmental mutation (environmental index is 0), the additional coefficient is zero, and the danger factor only reflects the independent risk to personnel.

[0044] Furthermore, when environmental indicators are zero (e.g., no sudden displacement in the foundation pit monitoring area), the risk factor depends entirely on the exceedance index, focusing on identifying the pressure trend of individual personnel. For example, if a steelworker's long-term head-down work causes a slow increase in pressure on their safety helmet but does not exceed the limit, an early warning can still be issued based on the continuous change trend, solving the problem of missing slow risks by static thresholds.

[0045] Furthermore, when the environmental indicator is 1 (such as a sudden change in infrared ranging caused by a broken bolt at a scaffold node), the additional coefficient becomes a risk "amplifier." This design is particularly suitable for indirect risk scenarios—for example, in a tower crane hoisting area, at the moment a steel cable breaks (a sudden environmental change) and causes a heavy object to fall, even if the pressure on the safety helmets of nearby workers is not exceeded (the exceedance index is low), the additional coefficient triggered by the sudden environmental change will significantly increase the danger factor in that area, thereby providing an early warning of the "range of the falling object shock wave." This is a cross-regional risk perception that isolated sensors cannot achieve.

[0046] In summary, by combining basic and additional coefficients, false alarms can be avoided (such as stable low-pressure states without environmental changes), and response sensitivity can be improved when multiple risks occur simultaneously, providing a high-confidence quantitative basis for predicting cross-regional behavior.

[0047] like Figure 6As shown, in one embodiment, S4, obtaining the i-th position change trend value of the construction worker to be identified based on the personnel location data sequence of the construction worker to be identified and the identification area where the i-th construction worker is located at the current moment, includes: S41. Obtain the relative distance value between each location data in the personnel location data sequence of the construction personnel to be identified and the identification area where the i-th construction personnel is located at the current time, and form a relative distance sequence of the construction personnel to be identified according to the order of the location data in the location data sequence. S42. Based on the difference between the next relative distance value and the previous relative distance value in the relative distance sequence of the construction personnel to be identified, the sum of all differences in the relative distance sequence is taken as the i-th position change trend value of the construction personnel to be identified.

[0048] In this embodiment, it should be noted that in S41, a spatiotemporal benchmark for cross-regional risk association is established. For each construction worker to be identified (e.g., worker B in a different area from worker A working in the high-risk tower crane area), their position sequence data for the previous time period is extracted (continuously recorded by a positioning wristband). Based on these discrete position points, the straight-line distance between worker B and the center point of the target identification area currently located by worker A at each moment is calculated, forming a "relative distance sequence" arranged in chronological order. For example, in the construction of a large factory, when the steel structure welding area (area X) is identified as a high-risk area due to the ignition of debris by an electric arc, the real-time distance between worker B in the pipeline installation area and area X within the previous 5 minutes is calculated, forming a sequence [5.2 meters → 4.8 meters → 4.3 meters from area X…]. This sequence eliminates absolute coordinate interference and focuses on expressing the positional change relationship between the target area and the person to be identified, providing spatiotemporal dimension input for subsequent trend analysis.

[0049] In S42, the incremental difference between adjacent distances (distance at the next moment minus distance at the previous moment) is calculated for the relative distance sequence generated in S41, and all differences in the sequence are summed as the position change trend value. This value reveals a continuous behavioral pattern: negative trend value: most sequence differences are negative (e.g., [-0.4, -0.5, -0.3]), indicating that personnel are continuously approaching the target area. For example, a maintenance worker traversing an equipment area to approach a leaking chemical tank, despite the winding path, the overall distance reduction trend is reinforced by the accumulated difference.

[0050] Furthermore, positive trend values: the differences are mostly positive (e.g., [+0.2, +0.6]), reflecting proactive avoidance behavior. It's important to note that this mechanism overcomes short-term fluctuations—if a person briefly approaches and then moves away (e.g., [-0.7, +1.2]), the positive difference will offset the negative value, avoiding misjudgment. This design breaks through the static nature of existing fence alarms, predicting the operational inertia of "moving towards a hazard source" based on historical trajectories.

[0051] like Figure 7As shown, in one embodiment, S4, obtaining the safety behavior identification result of the construction worker to be identified based on the trend value of the i-th location change of the construction worker to be identified and the danger coefficient of the identification area where the i-th construction worker is located at the current moment, includes: S43. If the trend value of the i-th position of the construction worker to be identified is less than zero, the product of the trend value of the i-th position of the construction worker to be identified and the danger coefficient of the identification area where the i-th construction worker is located at the current moment shall be used as the target identification value, and the behavior identification result of the construction worker to be identified shall be obtained based on the target identification value. S44. If the trend value of the i-th position of the construction worker to be identified is greater than zero, then the behavior identification result of the construction worker to be identified as a safe behavior is obtained.

[0052] In this implementation, it should be noted that in S43, the risk coupling between the high-risk environment and the approaching behavior is achieved. When the trend value in S42 is negative (person continuously approaches), it is multiplied by the risk coefficient calculated in S3 to generate the target identification value. This calculation reflects a dual risk amplification effect: behavioral risk weight: the degree of negativeness of the trend value (e.g., -8 compared to -3) reflects the speed and persistence of approach; the smaller the value (the larger the negative value), the stronger the risk-taking intention.

[0053] Furthermore, environmental risk weights: the risk factor comes from the coupling result of equipment stress and environmental abrupt changes within the region (such as tank leakage + maintenance worker's safety helmet stress).

[0054] For example, in tunnel collapse early warning, the danger factor of the collapse area increases due to rock displacement (sudden environmental changes) and excessive support pressure. If support workers approach the collapse point at this time (with a large negative trend value), the target identification value will drop sharply (e.g., -100 compared to -10), triggering the highest level alarm. This mechanism accurately quantifies the interaction effect between "proactive risk-taking" and "risk diffusion," avoiding the lag of existing alarms that only alert personnel who have already entered the danger zone.

[0055] In S44, scenarios without a risk-seeking intent are handled. When the trend value obtained from S42 is positive (personnel are generally moving away from the target area), the behavior is directly judged as safe. In dense construction scenarios, most personnel will naturally move during normal operations (such as concrete workers going around the tower crane area), and the positive trend value filters out such low-risk behaviors. Even if a sudden risk occurs in the target area (such as equipment tipping over), as long as the personnel to be identified do not actively approach, no warning is triggered to avoid excessive interference. For example, when the danger factor of the foundation pit support area suddenly increases, if the surrounding steelworkers continue to move away (trend value +4.2), even if they are only 3 meters away from the high-risk area, it is still marked as safe. This setting is consistent with engineering reality—risk diffusion has physical boundaries (such as the collapse radius), and no alarm is needed when personnel actively move away. Combined with the precise quantification of S43, an intelligent decision-making closed loop is achieved, ensuring that alarms are triggered when necessary and not triggered when not necessary.

[0056] A region-driven construction site safety behavior recognition system is also provided. In one embodiment, the system includes: The acquisition module is used to acquire the target construction site, which consists of multiple continuous identification areas. Based on the infrared ranging sensors deployed in each identification area, the pressure sensors in the safety helmets of construction workers, and the positioning wristbands of construction workers, it acquires the environmental change data sequence of each identification area in the previous acquisition time period, as well as the equipment status data sequence and personnel location data sequence of each construction worker. The personnel data processing module is used to obtain the i-th out-of-limit index based on the equipment status data sequence of the i-th construction worker, obtain the identification area where the i-th construction worker is located at the current moment, and obtain the i-th environmental change index based on the environmental change data sequence of the identification area where the i-th construction worker is located at the current moment. The regional data processing module is used to obtain the risk coefficient of the identification area where the i-th construction worker is located at the current moment based on the i-th environmental change index and the i-th exceedance index. The behavior recognition module is used to identify construction workers who are not in the same recognition area as the i-th construction worker at the current moment and to identify them as construction workers to be identified. Based on the personnel location data sequence of the construction workers to be identified and the recognition area where the i-th construction worker is located at the current moment, the module obtains the i-th position change trend value of the construction workers to be identified. Based on the i-th position change trend value of the construction workers to be identified and the danger coefficient of the recognition area where the i-th construction worker is located at the current moment, the module obtains the safety behavior recognition result of the construction workers to be identified.

[0057] In one embodiment, the personnel data processing module is further configured to: if there is a pressure value exceeding the pressure threshold in the equipment status data sequence of the i-th construction worker, then obtain the difference between the maximum pressure value and the pressure threshold and use it as the first over-limit value; obtain the change between the next pressure value and the previous pressure value in the equipment status data sequence of the i-th construction worker, and add all the changes in sequence to obtain the second over-limit value; if the second over-limit value is less than zero, then add the first over-limit value and zero to obtain the i-th over-limit index; if the second over-limit value is not less than zero, then add the first over-limit value and the second over-limit value to obtain the i-th over-limit index.

[0058] In one embodiment, the personnel data processing module is further configured to: determine whether there is fluctuation in the environmental change data sequence of the identification area where the i-th construction worker is located at the current moment, wherein the fluctuation is defined as the absolute change amplitude between adjacent distance data in the sequence exceeding a preset change threshold; if there is fluctuation, obtain a first preset value and use it as the i-th environmental change indicator; if there is no fluctuation, obtain a second preset value and use it as the i-th environmental change indicator.

[0059] In one implementation, the regional data processing module is further configured to: use the i-th exceedance index as a basic coefficient, use the product of the i-th exceedance index and the i-th environmental change index as an additional coefficient, and use the sum of the basic coefficient and the additional coefficient as the danger coefficient of the identification area where the i-th construction worker is located at the current moment.

[0060] In this embodiment, it should be noted that the specific method of performing the operation of the above-mentioned area-driven construction site safety behavior recognition system has been described in detail in the embodiments of the area-driven construction site safety behavior recognition method, and will not be elaborated here.

[0061] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.

[0062] It should also be noted that the various specific technical features described in the above embodiments can be combined in any suitable manner without contradiction. To avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.

[0063] Furthermore, various different embodiments of this disclosure can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content disclosed in this disclosure.

[0064] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A regional driving-based construction site safety behavior identification method, characterized in that, The method comprises the following steps: acquiring a target construction site composed of a plurality of continuous identification areas, and acquiring environmental change data sequences of the identification areas and equipment state data sequences and personnel position data sequences of the construction personnel in a previous collection time period before a current collection time period based on infrared distance sensors deployed in the identification areas, pressure sensors in safety helmets of the construction personnel, and positioning bands of the construction personnel; acquiring an i-th over-limit index according to the equipment state data sequence of the i-th construction personnel, and acquiring an identification area in which the i-th construction personnel is located at a current time, and acquiring an i-th environmental change index according to the environmental change data sequence of the identification area in which the i-th construction personnel is located at the current time; acquiring a danger coefficient of the identification area in which the i-th construction personnel is located at the current time according to the i-th environmental change index and the i-th over-limit index; acquiring construction personnel who are not in the same identification area as the i-th construction personnel at the current time as to-be-identified construction personnel, and acquiring an i-th position change trend value of the to-be-identified construction personnel according to the personnel position data sequence of the to-be-identified construction personnel and the identification area in which the i-th construction personnel is located at the current time, and acquiring a safety behavior identification result of the to-be-identified construction personnel according to the i-th position change trend value of the to-be-identified construction personnel and the danger coefficient of the identification area in which the i-th construction personnel is located at the current time.

2. The zone-driven construction site safety behavior recognition method according to claim 1, characterized in that, The method further comprises the following steps: if there is a pressure value exceeding a pressure threshold value in the equipment state data sequence of the i-th construction personnel, acquiring a difference between the maximum pressure value and the pressure threshold value as a first over-limit value; acquiring a change amount between a previous pressure value and a subsequent pressure value in the equipment state data sequence of the i-th construction personnel, and sequentially adding all the change amounts to obtain a second over-limit value; if the second over-limit value is less than zero, adding the first over-limit value and zero to obtain the i-th over-limit index, and if the second over-limit value is not less than zero, adding the first over-limit value and the second over-limit value to obtain the i-th over-limit index.

3. The zone-driven construction site safety behavior recognition method according to claim 1, characterized in that, The method further comprises the following steps: judging whether there is fluctuation in the environmental change data sequence of the identification area in which the i-th construction personnel is located at the current time, wherein the fluctuation is defined as an absolute change amplitude between adjacent distance data in the sequence exceeding a preset change threshold value; if there is fluctuation, acquiring a first preset value as the i-th environmental change index; if there is no fluctuation, acquiring a second preset value as the i-th environmental change index.

4. The zone-driven construction site safety behavior recognition method according to claim 1, characterized in that, The method further comprises the following steps: taking the i-th over-limit index as a basic coefficient, taking a product of the i-th over-limit index and the i-th environmental change index as an additional coefficient, and taking a sum of the basic coefficient and the additional coefficient as the danger coefficient of the identification area in which the i-th construction personnel is located at the current time.

5. The zone-driven construction site safety behavior recognition method according to claim 1, characterized in that, The i-th position change trend value of the to-be-identified construction worker is obtained according to the personnel position data sequence of the to-be-identified construction worker and the identification area in which the i-th construction worker is located at the current time, and the i-th position change trend value of the to-be-identified construction worker is obtained according to the personnel position data sequence of the to-be-identified construction worker and the identification area in which the i-th construction worker is located at the current time. The relative distance sequence of the to-be-identified construction worker is formed according to the relative distance values between each position data in the personnel position data sequence of the to-be-identified construction worker and the identification area in which the i-th construction worker is located at the current time in the order of the position data in the position data sequence. The i-th position change trend value of the to-be-identified construction worker is obtained according to the difference between the latter relative distance value and the former relative distance value in the relative distance sequence of the to-be-identified construction worker, and the sum of all the difference values in the relative distance sequence is taken as the i-th position change trend value of the to-be-identified construction worker.

6. The zone-driven construction site safety behavior recognition method according to claim 1, characterized in that, The safety behavior identification result of the to-be-identified construction worker is obtained according to the i-th position change trend value of the to-be-identified construction worker and the risk coefficient of the identification area in which the i-th construction worker is located at the current time, and the safety behavior identification result of the to-be-identified construction worker is obtained according to the i-th position change trend value of the to-be-identified construction worker and the risk coefficient of the identification area in which the i-th construction worker is located at the current time. The system comprises: The acquisition module is configured to acquire a target construction site composed of a plurality of continuous identification areas, and to acquire the environmental change data sequence of each identification area and the equipment state data sequence and the personnel position data sequence of each construction worker in a previous acquisition time period before a current acquisition time period based on the infrared distance measuring sensor deployed in each identification area, the pressure sensor in the safety helmet of the construction worker, and the positioning bracelet of the construction worker.

7. A zone-driven construction site safety behavior recognition system, characterized in that, The personnel data processing module is configured to acquire the i-th over-limit index according to the equipment state data sequence of the i-th construction worker, to acquire the identification area in which the i-th construction worker is located at the current time, and to acquire the i-th environmental change index according to the environmental change data sequence of the identification area in which the i-th construction worker is located at the current time. The area data processing module is configured to acquire the risk coefficient of the identification area in which the i-th construction worker is located at the current time according to the i-th environmental change index and the i-th over-limit index. The behavior identification module is configured to acquire, at the current time, a construction worker who is not in the same identification area as the i-th construction worker as a to-be-identified construction worker, to acquire the i-th position change trend value of the to-be-identified construction worker according to the personnel position data sequence of the to-be-identified construction worker and the identification area in which the i-th construction worker is located at the current time, and to acquire the safety behavior identification result of the to-be-identified construction worker according to the i-th position change trend value of the to-be-identified construction worker and the risk coefficient of the identification area in which the i-th construction worker is located at the current time. The personnel data processing module is further configured to: If there is a pressure value exceeding the pressure threshold value in the equipment state data sequence of the i-th construction worker, the difference between the maximum pressure value and the pressure threshold value is taken as the first over-limit value.

8. The zone-driven construction site safety behavior recognition system of claim 7, wherein, ​ ​ obtain a change between a latter pressure value and a former pressure value in the equipment state data sequence of the i th construction worker, add all the changes in turn and obtain a second overrun value; if the second overrun value is less than zero, add the first overrun value and zero to obtain the i th overrun index, and if the second overrun value is not less than zero, add the first overrun value and the second overrun value to obtain the i th overrun index.

9. The zone-driven construction site safety behavior recognition system of claim 7, wherein, The personnel data processing module is further configured to: determine whether there is fluctuation in the environment change data sequence of the identified area where the i th construction worker is located at the current time, wherein the fluctuation is defined as an absolute change amplitude between adjacent distance data in the sequence exceeding a preset change threshold; if there is fluctuation, obtain a first preset value as the i th environment change index; if there is no fluctuation, obtain a second preset value as the i th environment change index.

10. The zone-driven construction site safety behavior recognition system of claim 7, wherein, The area data processing module is further configured to: take the i th overrun index as a basic coefficient, take a product of the i th overrun index and the i th environment change index as an additional coefficient, and take a sum of the basic coefficient and the additional coefficient as a danger coefficient of the identified area where the i th construction worker is located at the current time.

Images