A labor scheduling system for construction subcontracting

Abstract

The application discloses a laborer dispatching system for building labor subcontracting, and particularly relates to the technical field of labor resource dispatching. A basic information module stores personnel information; a real-time sensing module collects real-time data; a task decomposition module decomposes an engineering into operation units; an intelligent matching module preliminarily screens to form a candidate set; a dynamic regulation and control module, when there is a resource conflict, calculates a risk index, an influence coefficient and a coupling degree based on physiological parameters, a progress plan and a process logic, obtains an adaptation degree through fuzzy logic, and globally optimizes to generate a dispatching scheme; a guidance intervention module pushes information to a person in charge when there is no feasible scheme; the application fuses real-time sensing and a building information model, and realizes safe, progress and quality collaborative balance through multi-dimensional index calculation, fuzzy logic and global optimization, so that a decision-making stalemate of losing one to save the other is avoided; meanwhile, system reliability under extreme conditions is ensured through a man-machine collaborative fault-tolerant mechanism.

Description

Technical Field

[0001] This invention relates to the field of labor resource scheduling technology, and more specifically, to a labor dispatch system for construction labor subcontracting. Background Technology

[0002] With the advancement of digital transformation in the construction industry, labor dispatch and management are gradually evolving from traditional manual scheduling to digitalization and intelligence. Various existing labor resource management systems manage basic labor information through real-name registration, skill tags, and attendance records. They can also integrate construction schedules from building information models to break down project tasks into specific work units, achieving initial personnel matching based on skill requirements.

[0003] In actual engineering projects, multiple work units often start work simultaneously within the same time period, creating competition for limited labor resources. When the same pool of candidates cannot simultaneously meet the labor needs of multiple work units, resource conflicts arise. Existing technologies typically address such conflicts by using preset priority rules or simple manual intervention. For example, assigning work units sequentially according to their planned start times, or making temporary adjustments based on the subjective experience of managers.

[0004] However, conventional scheduling methods face a dilemma under multiple constraints, making it difficult to balance all aspects. On the one hand, prioritizing work units on the critical path may lead to the hasty assignment of workers in poor condition or with low skill matching to other work units, creating safety hazards or quality risks. On the other hand, strictly screening workers to pursue quality may cause delays in key processes, affecting the overall project schedule. This trade-off between multiple objectives such as schedule, safety, and quality creates a prisoner's dilemma-like stalemate in scheduling decisions, making it difficult to achieve a balance among various constraints regardless of the assigned work plan. Therefore, this invention proposes a labor scheduling system for construction labor subcontracting to solve the above problems. Summary of the Invention

[0005] To achieve the above objectives, the present invention provides the following technical solution: A labor dispatch system for construction labor subcontracting includes: The basic information module is used to store and manage the identity information, skill qualifications, and labor contract information of the workers. The real-time sensing module is used to collect the real-time location, physiological parameters, and environmental data of the workers' work site through wearable terminal devices. The task decomposition module is used to decompose engineering tasks into work units that include corresponding skill requirements, time windows, and quality requirements based on the construction schedule imported from the preset building information model. The intelligent matching module is used to perform preliminary matching based on the skill requirements of the work unit and the skill qualifications of the workers, and to filter out the workers who are in an idle state by combining real-time location to form a candidate set. The dynamic control module is used to calculate the worker status risk index based on the physiological status parameters collected by the real-time perception module when the candidate personnel set output by the intelligent matching module cannot simultaneously meet the priority requirements of multiple work units and there are resource conflicts. Based on the schedule plan and process logic in the building information model, it calculates the critical path influence coefficient of the task and the process quality coupling degree. Taking the worker status risk index, the critical path influence coefficient of the task and the process quality coupling degree as input parameters, it calculates the suitability of each candidate worker for the conflicting work units through fuzzy logic rules, and performs global combination optimization based on the suitability calculation results to generate the optimal work assignment plan that meets the safety bottom line, schedule constraints and quality requirements. The guidance and intervention module is used to push information from the task decomposition module, intelligent matching module, and dynamic control module to the project leader when safety bottom lines, schedule constraints, and quality requirements cannot be met, serving as the basis for their actual guidance and intervention.

[0006] In a preferred embodiment, the identity information includes the worker's historical health records and age data, the skill qualifications include the worker's labor skills and corresponding levels, and the labor contract information includes the worker's historical performance data and historical quality traceability data.

[0007] In a preferred embodiment, the preset building information model is a structured database containing three-dimensional component geometric data, process logic relationships, planned start time, planned completion time, and preset quality acceptance standards.

[0008] In a preferred embodiment, decomposing an engineering task into work units that include corresponding skill requirements, time windows, and quality requirements means: Based on the process logic in the building information model, the unit project is divided into several construction sections. The required labor skills and corresponding levels are matched from the preset skill library according to the component type corresponding to the construction section to form skill requirements. The workable time interval is determined by combining the planned start time and planned completion time with the process logic to form a time window. The corresponding process requirements and acceptance indicators are extracted from the preset quality library according to the preset quality acceptance standards to form quality requirements.

[0009] In a preferred embodiment, in the dynamic control module, the worker status risk index is calculated by combining the physiological status parameters collected by the real-time sensing module and the environmental data of the work site with the historical health records and age data of the corresponding laborers stored in the basic information module. The critical path impact coefficient is calculated based on the schedule in the building information model, combined with the historical performance data contained in the labor contract information of the corresponding laborers stored in the basic information module, and the time taken to reach the work surface reflected by the real-time location collected by the real-time sensing module. The process quality coupling degree is obtained by correlation analysis based on the process logic in the building information model, combined with the historical quality traceability data contained in the labor contract information of the corresponding laborers stored in the basic information module, and the degree of conformity of the work surface and the degree of environmental suitability reflected by the real-time location and environmental data collected by the real-time sensing module.

[0010] In a preferred embodiment, the worker condition risk index is calculated according to the following steps: Based on the parameters collected by the real-time sensing module, and according to the preset baseline interval and safety threshold, the measured values ​​are mapped to single risk factors in the [0,1] interval using a piecewise linear function: if the parameter has a baseline interval, the risk factor is 0 when the measured value is within the baseline interval; when it exceeds the interval but does not exceed the safety threshold, the risk factor is proportional to the degree of exceeding; when it exceeds the safety threshold, it is always 1; if the parameter only has a safety threshold, the risk factor is proportional to the measured value when the measured value is below the safety threshold; when it exceeds the safety threshold, it is always 1; the physiological load index is obtained by weighted summation of the risk factors of heart rate and body temperature; the environmental stress index is obtained by weighted summation of the risk factors of temperature, humidity, and dust. Based on the age stored in the basic information module, the age correction coefficient is obtained according to the preset age-risk coefficient mapping table; Based on the historical health records stored in the basic information module, the health correction coefficient is determined according to the number of basic disease types recorded therein; the more disease types there are, the larger the health correction coefficient will be. The physiological stress index and the environmental stress index are added together, and then multiplied by the product of the age correction factor and the health correction factor to obtain the worker's condition risk index.

[0011] In a preferred embodiment, the critical path impact coefficient of the task is calculated according to the following steps: Based on the schedule in the building information model, the planned start time and planned completion time of the current work unit are extracted, the total float time of the work unit is calculated, and the foundation influence coefficient is determined linearly according to the length of the total float time. The shorter the total float time, the larger the foundation influence coefficient. Based on the historical performance data contained in the labor contract information of the corresponding laborers stored in the basic information module, the ratio of the laborer's average work efficiency to the standard work efficiency is calculated and used as the efficiency correction coefficient. Based on the real-time location collected by the real-time sensing module, the path distance between the worker's current location and the work surface of the conflicting work unit is calculated. The arrival time is estimated according to the preset moving speed, and the time delay correction coefficient is determined based on the difference between the arrival time and the planned start time. Multiply the base impact coefficient by the efficiency correction coefficient, and then multiply by the time delay correction coefficient to obtain the critical path impact coefficient of the mission.

[0012] In a preferred embodiment, the process quality coupling degree is calculated according to the following steps: Based on the process logic in the building information model, the number of subsequent processes for the current work unit is determined, and the unit rework cost of the corresponding component type for the current work unit is obtained from the preset rework cost library. The product of the number of subsequent processes and the unit rework cost is divided by the preset benchmark value to obtain the process dependence strength coefficient. Based on the historical quality traceability data of the corresponding laborers stored in the basic information module, the pass rate of the laborer's work records with the same component type as the current work unit is statistically analyzed within a preset time window, and used as the skill reliability coefficient. Based on the real-time location collected by the real-time sensing module, the distance between the current location of the worker and the center point of the work surface corresponding to the current work unit in the building information model is calculated. The distance is converted into a location conformity coefficient in the range of [0,1] according to the preset distance-conformity mapping function. The distance-conformity mapping function is as follows: when the distance is less than or equal to the first threshold, the location conformity coefficient is 1; when the distance is greater than or equal to the second threshold, the location conformity coefficient is 0; when the distance is between the first threshold and the second threshold, the location conformity coefficient decreases linearly with the increase of distance. Based on the environmental data collected by the real-time sensing module, the deviation values ​​between the current ambient temperature, ambient humidity, and ambient wind speed and the standard values ​​of the process environment requirements corresponding to the current work unit in the building information model are calculated respectively. After normalizing each deviation value, the maximum value is taken as the environmental deviation index. The environmental compliance coefficient is obtained by subtracting the environmental deviation index from 1. The environmental compliance coefficient is truncated to the interval [0,1]. The process quality coupling degree is obtained by multiplying the process dependence intensity coefficient, skill reliability coefficient, location conformity coefficient, and environment conformity coefficient.

[0013] In a preferred embodiment, the suitability of each candidate worker for the conflicting work unit is calculated using fuzzy logic rules, specifically as follows: The worker status risk index, the critical path influence coefficient of the task, and the process quality coupling degree are used as the three input variables of the fuzzy logic system. The precise values ​​of each input variable are converted into corresponding fuzzy values ​​according to the preset membership function. The fuzzy values ​​include three linguistic values: low, medium, and high. The fuzzy values ​​of the three input variables are input into a preset fuzzy rule base for fuzzy inference. The inference rules output corresponding fuzzy values ​​of fit degree based on different combinations of worker state risk index, task critical path influence coefficient and process quality coupling degree. The fuzzy values ​​of fit degree include five linguistic values: low, lower, medium, higher and high. The fuzzy value of the fit is converted into a precise value in the range [0,1] using a preset defuzzification method, so as to obtain the fit of the candidate worker with respect to the current conflicting work unit.

[0014] In a preferred embodiment, global combination optimization is performed based on the results of the fit calculation to generate an optimal work assignment plan that meets the safety baseline, schedule constraints, and quality requirements, specifically as follows: Construct a fit degree matrix with multiple work units currently in conflict as rows and multiple laborers in the candidate set as columns. Each element in the fit degree matrix represents the fit degree of the corresponding laborer for the corresponding work unit. With the objective function of maximizing the sum of the fit of all dispatched workers to their assigned work units, and with safety baseline constraints, schedule constraints, and quality requirements as constraints, an integer programming model is constructed. The safety baseline constraint is that the worker status risk index of each worker must be lower than the preset risk threshold. The schedule constraint is that each work unit must be completed within its time window and the sum of the arrival times of all dispatched workers must not exceed the upper limit of the delay allowed by the planned start time. The quality requirement is that the process quality coupling degree of the dispatched workers in each work unit must be higher than the preset quality threshold. Solve the integer programming model to obtain the optimal matching relationship between each conflicting work unit and the assigned workers, as well as the list of unassigned workers.

[0015] The technical effects and advantages of this invention are as follows: This invention achieves a dynamic balance under multiple constraints of safety, schedule, and quality, effectively breaking the prisoner's dilemma-style decision-making deadlock when resources conflict. Existing technologies often suffer from imbalances when multiple work units compete for the same batch of workers. Prioritizing critical path operations may neglect worker physical condition and environmental risks, while prioritizing construction quality may lead to project delays. This invention quantifies real-time physiological parameters and environmental stress levels through a worker condition risk index, accurately reflects the impact of personnel allocation on the overall project schedule through a critical path impact coefficient, and comprehensively measures the potential impact of historical quality performance, real-time location, and environmental suitability on process quality through process quality coupling degree. Using these three dimensions as input parameters for fuzzy logic rules, the invention calculates the suitability of each candidate worker for conflicting work units. Then, through global combinatorial optimization, it solves for the optimal matching relationship that satisfies safety baselines, schedule constraints, and quality requirements. This achieves synergistic optimization among the three objectives of schedule, safety, and quality, avoiding systemic imbalances caused by simple prioritization or subjective experience-based decisions.

[0016] This invention deeply couples real-time sensing data with Building Information Modeling (BIM), enabling scheduling decisions to dynamically respond to changes on-site. Existing scheduling systems often rely on static skill tags and fixed construction plans, failing to perceive dynamic fluctuations in worker location, heart rate, body temperature, and environmental factors such as work surface temperature, humidity, and dust. This invention continuously collects this dynamic data through a real-time sensing module and integrates it into correction calculations for worker status risk indices and process quality coupling. For example, it estimates arrival time based on real-time location to correct the critical path impact coefficient, and assesses compliance with process requirements based on ambient temperature and humidity to correct the process quality coupling. This allows the work assignment scheme to respond to immediate changes in worker fatigue levels and environmental severity, avoiding the issuance of risky work instructions due to static data lagging behind actual on-site conditions, thus reducing the probability of safety accidents and quality defects at the source.

[0017] This invention constructs a two-layer fault-tolerant mechanism combining automated algorithmic decision-making and manual guidance to ensure system robustness under extreme conditions. Existing technologies typically report errors or interrupt processes directly when the algorithm has no solution, relying on temporary on-site coordination by management personnel, which lacks information support and leads to blind decision-making. This invention sets up a guidance and intervention module. When the dynamic control module, after global optimization, still cannot find a feasible solution that simultaneously meets safety baselines, schedule constraints, and quality requirements, it automatically pushes task decomposition information, candidate personnel sets, various risk indicators, suitability calculation results, and specific violated constraints to the project manager's terminal, presenting the full picture of the conflict in a visual manner. This allows the manager to make adjustment decisions based on comprehensive data support and on-site experience, such as relaxing the schedule tolerance for non-critical paths, urgently transferring personnel from other sections, or temporarily adjusting the environmental conditions of the work surface. This fills the decision-making blind spot of the fully automated scheduling system under extremely complex conditions, significantly improving the system's adaptability and reliability in practical engineering applications. Attached Figure Description

[0018] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings; Figure 1 This is a schematic diagram of a labor dispatch system for construction labor subcontracting according to the present invention. Detailed Implementation

[0019] 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. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0020] Reference Figure 1 The following examples were obtained: Example 1: A labor dispatch system for construction labor subcontracting, comprising: The basic information module is used to store and manage the identity information, skill qualifications, and labor contract information of laborers. By centrally storing the identity information (such as age and health records), skill qualifications (such as job type and skill level), and labor contract information (such as historical performance and quality traceability data) of laborers, it provides static data support for subsequent intelligent matching and dynamic control, ensuring that the system can accurately identify the basic attributes and historical performance of each laborer.

[0021] The real-time sensing module is used to collect real-time location, physiological parameters, and environmental data of the work site of workers through wearable terminal devices. This module is the sensing layer of the system. Through terminal devices such as smart safety helmets and wristbands, it can obtain real-time location dynamics, physiological indicators such as heart rate and body temperature of workers, as well as environmental parameters such as temperature, humidity, and dust of the work site. It provides the system with dynamic, on-site real-time data, enabling dispatch decisions to respond to the actual situation on site.

[0022] The task decomposition module is used to decompose engineering tasks into work units containing corresponding skill requirements, time windows, and quality requirements based on the construction schedule imported from the preset building information model. This module transforms the macro-level construction plan into executable micro-level task units. By analyzing the process logic, time nodes, and quality standards in the building information model, it clarifies the skills required for each work unit, the allowed time range for operation, and the quality requirements that must be met, thus laying the foundation for accurate work assignment.

[0023] The intelligent matching module is used to perform an initial match between the skill requirements of the work unit and the skill qualifications of the workers, and to filter out the workers who are in an idle state by combining real-time location, forming a candidate pool. This module performs the first layer of screening, first ensuring that the skills of the workers meet the task requirements, and then excluding the workers who are already on duty by real-time location, generating a candidate pool that meets the basic conditions, so as to narrow down the decision-making scope for subsequent dynamic optimization.

[0024] The dynamic control module is used when the candidate set output by the intelligent matching module cannot simultaneously meet the priority requirements of multiple work units and resource conflicts exist. It calculates the worker status risk index based on the physiological state parameters collected by the real-time perception module, and calculates the critical path impact coefficient and process quality coupling degree based on the schedule plan and process logic in the building information model. Taking the worker status risk index, critical path impact coefficient, and process quality coupling degree as input parameters, it calculates the suitability of each candidate worker for the conflicting work units through fuzzy logic rules, and performs global combination optimization based on the suitability calculation results to generate the optimal work assignment plan that meets the safety baseline, schedule constraints, and quality requirements. This module is the core intelligent layer of the system. It introduces multi-dimensional dynamic indicators when resource conflicts occur, handles uncertainty through fuzzy logic, and finally achieves multi-objective balance through combination optimization to ensure that the work assignment plan achieves the optimal solution between safety, schedule, and quality.

[0025] The guidance and intervention module is used to push information from the task decomposition module, intelligent matching module, and dynamic control module to the project leader when safety baselines, schedule constraints, and quality requirements cannot be met, serving as the basis for their actual guidance and intervention. This module also acts as an interface for human-machine collaboration. When the algorithm cannot automatically generate a feasible solution, it summarizes and pushes information such as conflict details, candidate personnel status, and task requirements to the manager, allowing human experts to make the final decision based on their experience, ensuring the system's robustness in extreme situations.

[0026] In one specific implementation, taking a commercial complex project as an example, the composition of basic information is explained in detail. First, the system establishes an independent data file for each worker entering the site. This file includes three components: identity information, skill qualifications, and labor contract information. Specifically, the identity information includes the worker's historical health records and age data. For example, for steelworker Zhang San, his identity information stores his historical health records, which include his physical examination results over the past three years, including a blood pressure of 125 mmHg in 2021, 130 mmHg in 2022, and a new diagnosis of mild fatty liver in 2023. His age is also recorded as 45 years old. For welder Li Si, his identity information stores his historical health records, indicating no history of underlying diseases, a visual acuity of 5.0 in both eyes, and an age of 32 years old. The skill qualifications in the identity information specifically include the worker's labor skills and corresponding levels. For example, Zhang San's skill qualification record shows that his skill category is steel bar worker, the corresponding level is senior worker, and the validity period is until 2026; Li Si's skill qualification record shows that his skill category is welder, the corresponding level is intermediate worker, and he also holds a special operation certificate, which allows him to operate fusion welding and thermal cutting operations.

[0027] The employment contract information within the identity information specifically includes historical performance data and historical quality traceability data for the workers. For example, Zhang San's employment contract information records that his average efficiency in completing standard floor rebar tying work in previous projects was 1.2 tons per workday, exceeding the quota efficiency by 20%. His historical quality traceability data records that in the five projects he participated in, the first-time acceptance pass rate was 96%, and includes the original record numbers for each acceptance, including the acceptance record of the fifth-floor beams and slabs of a residential building project showing that the rebar spacing pass rate in his area was 98%. Similarly, Li Si's employment contract information records that in a steel structure factory project, he completed a total weld length of 250 meters, averaging 8.3 meters of welds per day. His historical quality traceability data records that his first-time pass rate for ultrasonic flaw detection was 92%, and includes detailed descriptions of three rework records, including the reason for rework, rework time, and re-inspection results. The aforementioned identity information, skill qualifications, and employment contract information together constitute a basic database, providing static data support for all subsequent scheduling decisions.

[0028] In one specific implementation, considering a scenario where resource conflicts arise between two work units—the core tube concrete pouring and the installation of embedded parts for the outer curtain wall—in a large commercial complex project, the labor dispatch method is described in detail. First, a pre-set Building Information Model (BIM) is retrieved. This model is a structured database containing three-dimensional component geometric data, process logic relationships, planned start times, planned completion times, and pre-set quality acceptance standards. Decomposing the project task into work units containing corresponding skill requirements, time windows, and quality requirements involves: dividing the unit project into several construction sections based on the process logic relationships in the BIM; matching the required labor skills and corresponding levels from a pre-set skill library based on the component types corresponding to each construction section to form skill requirements; determining the workable time interval based on the planned start and completion times combined with the process logic relationships to form time windows; and extracting the corresponding process requirements and acceptance indicators from a pre-set quality library based on the pre-set quality acceptance standards to form quality requirements.

[0029] A specific example is the division of a unit project into several construction sections. For instance, for the three-story underground structure of this project, the model shows that the floor plan is divided into three sections: Area A, Area B, and Area C. Area A contains eight frame columns, six shear walls, and one floor slab. There are clear construction sequence constraints between the components: the formwork can only be installed after the column and wall reinforcement is tied, and the concrete can only be poured after the formwork is installed. Based on the above process logic, the system further subdivides Area A into five basic work units: column and wall reinforcement tying, column and wall formwork installation, floor slab formwork installation, column and wall concrete pouring, and floor slab concrete pouring. Subsequently, the system matches the required labor skills and corresponding levels from a preset skill library according to the component type corresponding to each construction section to form skill requirements. Taking the column and wall reinforcement binding operation unit in Area A as an example, the system identifies its component type as frame column and shear wall, the reinforcement specifications are mainly 25 mm and 20 mm in diameter, and the node structure is complex. Therefore, the skill requirement matched from the preset skill library is a steelworker with an advanced certificate. As for the floor slab formwork installation operation unit, its component type is beam-slab system, and the formwork installation difficulty is relatively low. The skill requirement matched by the system is a carpenter with an intermediate certificate.

[0030] The system determines the workable time intervals to form time windows based on the planned start and finish times combined with the logical relationships of the work processes. For example, the building information model records that the planned start time for the column and wall reinforcement binding in Area A is 8:00 AM on March 15th, and the planned finish time is 5:00 PM on March 17th. Furthermore, the subsequent column and wall formwork installation can only begin after the reinforcement binding is completed. Therefore, the system determines the time window for this work unit to be from 8:00 AM on March 15th to 5:00 PM on March 17th. However, for the floor slab formwork installation in Area A, since it is logically independent of the column and wall work, and the model shows that its planned start time is 8:00 AM on March 16th, and its planned finish time is 5:00 PM on March 18th, its time window is determined to be from 8:00 AM on March 16th to 5:00 PM on March 18th.

[0031] The system extracts corresponding process requirements and acceptance indicators from a pre-set quality database based on preset quality acceptance standards to form quality requirements. Taking the column and wall reinforcement binding operation unit as an example, the preset quality acceptance standard associated with the building information model is the current national standard for construction quality acceptance of concrete structures. The system extracts corresponding process requirements from the preset quality database, including specific parameters such as reinforcement specifications, spacing, protective layer thickness, and anchorage length, as well as acceptance indicators such as allowable deviations of ±10 mm for the spacing of stressed reinforcement, ±20 mm for the spacing of stirrups, and ±5 mm for the thickness of the protective layer. For the concrete pouring operation unit, the system extracts quality requirements including process requirements such as slump control between 180 mm and 220 mm, vibration spacing not exceeding 500 mm, and no cold joints, as well as acceptance indicators such as axial position deviation not exceeding 8 mm and cross-sectional dimension deviation between +5 mm and -5 mm. Through the above process, the system transforms the macro-level construction plan into a series of micro-level operation units with clear skill requirements, strict time windows, and specific quality requirements, providing precise task basis for subsequent personnel matching and scheduling. After completing the above task decomposition, the system enters the intelligent matching phase. Based on the skill requirements of the work units and the skill qualifications of the workers, a preliminary screening is performed. For example, all workers holding advanced concrete worker certificates and with excellent past performance are retrieved from the database. Combined with real-time location data collected by the real-time sensing module, workers already resting on other work surfaces or in the living area are excluded, ultimately forming a candidate pool. At this point, the system detects that only five qualified workers are available for deployment for the critical path operation of core tube pouring, while the installation of peripheral embedded parts also requires three of these five workers for precise layout, thus creating a resource conflict. To resolve this conflict, the system enters the dynamic control phase.

[0032] In the dynamic control module, the worker status risk index is calculated by combining the physiological status parameters collected by the real-time sensing module and the environmental data of the work site with the historical health records and age data of the corresponding workers stored in the basic information module. The critical path impact coefficient is calculated based on the schedule in the building information model, combined with the historical performance data contained in the labor contract information of the corresponding laborers stored in the basic information module, and the time taken to reach the work surface reflected by the real-time location collected by the real-time sensing module. The process quality coupling degree is obtained by correlation analysis based on the process logic in the building information model, combined with the historical quality traceability data contained in the labor contract information of the corresponding laborers stored in the basic information module, and the degree of conformity of the work surface and the degree of environmental suitability reflected by the real-time location and environmental data collected by the real-time sensing module.

[0033] In one specific implementation, taking the real-time status assessment of worker Zhang San during the core tube construction of a super high-rise project as an example, the calculation process of the worker status risk index is explained in detail. First, the system collects Zhang San's physiological status parameters and environmental data of the work site in real time through a wearable terminal device. At this time, the collected heart rate is 96 beats per minute, body temperature is 37.1 degrees Celsius, the ambient temperature of the work site is 33 degrees Celsius, the ambient humidity is 78%, and the ambient dust concentration is 0.35 milligrams per cubic meter. Based on the above five parameters, according to preset benchmark intervals and safety thresholds, the system maps each measured value to a single risk factor within the [0,1] interval using a piecewise linear function. For heart rate parameters, the preset normal range is 60 to 80 beats per minute. The safety threshold consists of a lower safety threshold and an upper safety threshold. The lower safety threshold is 50 beats per minute and the upper safety threshold is 120 beats per minute. Since the actual measured value of 96 beats per minute exceeds the upper limit of the normal range, the system calculates the proportion of the excess part to the distance between the boundary of the normal range and the safety threshold, that is, (96-80) / (120-80)=0.4, and the heart rate risk factor is 0.4.

[0034] For body temperature parameters, the preset normal range is 36.2°C to 37.2°C, the lower safety threshold is 35.5°C, and the upper safety threshold is 38.5°C. The measured value of 37.1°C is within the normal range, therefore the body temperature risk factor is 0. For ambient temperature parameters, the preset comfort range is 18°C ​​to 26°C, and the upper safety threshold is 36°C. The measured value of 33°C exceeds the comfort range. Based on the ratio of the excess (33-26) to the distance between the comfort range boundary and the safety threshold (36-26) being 0.7, the temperature risk factor is calculated to be 0.7. For the environmental humidity parameter, the preset comfort range is 40% to 60%, and the safe upper limit threshold is 85%. The measured value is 78%, exceeding the comfort range. Calculated using (78-60) / (85-60) = 0.72, the humidity risk factor is 0.72. For the environmental dust concentration parameter, this parameter only has a safe threshold and no comfort range. The preset safe threshold, i.e., the limited safe value, is 4 mg / m³. The measured value is 0.35 mg / m³, which is below the limit. Therefore, calculated using the ratio of the measured value to the safe threshold, 0.35 / 4 = 0.0875, the dust risk factor is 0.0875. After completing the mapping of the above individual risk factors, the system enters the synthesis stage of the physiological load index and the environmental stress index.

[0035] In existing technologies, the assessment of multi-parameter health risks typically employs equal-weighted averaging or expert scoring. In addition to the aforementioned existing technologies, this specific implementation can also adopt a differentiated weighting setting based on the combined effects of physiological sensitivity and environment. The weighting of heart rate and body temperature is based on the fact that heart rate fluctuations have a greater impact on the body's immediate working capacity than slight fluctuations in body temperature. Medical research shows that heart rate changes directly reflect the cardiovascular system load, while the body still has a relatively strong ability to regulate core body temperature even with slight fluctuations. Therefore, the weight of heart rate is set to 0.7 and the weight of body temperature is set to 0.3. The physiological load index of 0.28 is obtained by weighted summation of the heart rate risk factor of 0.4 and the body temperature risk factor of 0. For environmental parameters, temperature and humidity have a significant coupling effect on heat stress. According to the human thermal comfort equation and occupational health standards, the contribution rate of temperature to heat stress is about 60%, humidity is about 30%, and dust is about 10%. Therefore, after multiplying the temperature risk factor 0.7, humidity risk factor 0.72, and dust risk factor 0.0875 by their corresponding weights of 0.6, 0.3, and 0.1 respectively and summing them, we get the environmental stress index 0.7×0.6+0.72×0.3+0.0875×0.1=0.42+0.216+0.00875=0.64475, or about 0.645.

[0036] The system retrieves Zhang San's identity information from the basic database. The historical health records and age data show that Zhang San is currently 48 years old, and his health records indicate he has two underlying conditions: grade 2 hypertension and mild diabetes. For determining the age correction factor, the system uses a pre-defined age-risk coefficient mapping table. This table, constructed based on large-sample occupational health statistics, divides age into four ranges: 18-35, 36-45, 46-55, and 56+, with corresponding correction factors of 1.0, 1.1, 1.3, and 1.6, respectively. Since Zhang San is 48 years old, falling within the 46-55 range, the age correction factor is set to 1.3. The system calculates the health correction coefficient based on the number of underlying diseases recorded in historical health records. The theoretical basis for this is that the coexistence of multiple diseases leads to a decline in the body's physiological reserve function and a weakening of stress compensation ability. In medical research, the number of comorbidities is often used as an important predictive indicator of health risk. In this specific implementation, the health correction coefficient is set to 1.0 when the number of underlying diseases is 0. The coefficient increases by 0.2 for each additional disease, and accumulates to a maximum of 1.8. Zhang San suffers from both hypertension and diabetes, so the health correction coefficient is 1.0 + 0.2 × 2 = 1.4. Finally, the system adds the physiological load index (0.28) and the environmental stress index (0.645) to obtain the median value of 0.925. Then, it multiplies this value by the product of the age correction factor (1.3) and the health correction factor (1.4), which is 1.82. The final result is Zhang San's current worker status risk index, which is 0.925 × 1.82 = 1.6835. This value exceeds the preset risk concern threshold of 1.5. Based on this, the system avoids assigning Zhang San to high-altitude operations or physically demanding positions in subsequent scheduling, thereby achieving risk warning and control based on the fusion of real-time status and basic information.

[0037] In one specific implementation, taking the shear wall reinforcement binding operation in the core tube area of ​​a super high-rise building as an example, the calculation process of the critical path influence coefficient is explained in detail. The planned start time and planned finish time of the current work unit are extracted from the Building Information Model (BIM). The total float time of the work unit is calculated, and the foundation influence coefficient is linearly determined based on the length of the total float time; the shorter the total float time, the larger the foundation influence coefficient. For example, the shear wall reinforcement binding operation unit on the seventh floor of the core tube is shown in the schedule as having a planned start time of 8:00 AM and a planned finish time of 5:00 PM on the same day. By analyzing that the planned start time of the subsequent process, formwork installation, is 8:00 AM the following day, the total float time of this work unit can be calculated to be 15 hours. Since the total float time is relatively long, it indicates that the work unit has a higher tolerance for schedule delays, and therefore the foundation influence coefficient is relatively small. Conversely, if the total float time of another core tube concrete pouring operation unit is zero and it is on the critical path, then its foundation influence coefficient takes the maximum value of 1.0. In this implementation, the total floating time is set to have a negative linear relationship with the basic influence coefficient. For example, when the total floating time is less than or equal to two hours, the basic influence coefficient is 1.0; when it is greater than or equal to twenty-four hours, the basic influence coefficient is 0.2. The intermediate values ​​are calculated by linear interpolation. In this example, the basic influence coefficient corresponding to fifteen hours is calculated to be 0.45.

[0038] Based on the historical performance data of the corresponding workers stored in the labor contract information, the ratio of the worker's average work efficiency to the standard work efficiency is calculated as the efficiency correction coefficient. Taking steelworker Li Si as an example, his historical performance data in his labor contract information shows that in four similar shear wall steel reinforcement binding operations he participated in over the past six months, his average daily steel reinforcement binding volume was 1.8 tons, while the industry standard work efficiency is 1.5 tons per day. The ratio of his average work efficiency to the standard work efficiency is 1.8 divided by 1.5, which equals 1.2, meaning the efficiency correction coefficient is 1.2. This indicates that the worker's work efficiency is above average and can have a positive impact on the project schedule. If another worker, Li Ming, has an average work efficiency of only 1.2 tons, then his efficiency correction coefficient is 0.8, reflecting that his work efficiency is low and may increase the risk of delays.

[0039] Based on the real-time location data collected by the real-time sensing module, the path distance between the worker's current location and the work surface of the conflicting work unit is calculated. The arrival time is estimated based on a preset movement speed, and a time delay correction coefficient is determined based on the difference between the arrival time and the planned start time. Taking Li Si as an example, the real-time sensing module detects that he is currently located in the dormitory building in the living area. The shortest path length from his coordinates to the seventh-floor work surface of the core tube is 350 meters. Based on the preset walking speed of 1.2 meters per second on the construction site, his arrival time is estimated to be 350 divided by 1.2, approximately 292 seconds, or about 4.9 minutes. The current time is 7:55 AM, the planned start time is 8:00 AM, and the arrival time is approximately 8:05 AM, which is 5 minutes later than the planned start time. According to the preset time delay correction function, the correction coefficient increases linearly when the delay is within five minutes; for example, the correction coefficient is 1.01 for a one-minute delay and 1.05 for a five-minute delay. If the delay exceeds 30 minutes, the upper limit may be applied directly. In this embodiment, a five-minute delay corresponds to a time delay correction coefficient of 1.05. If workers arrive early, the correction factor can be less than one, indicating that arriving early is beneficial to the project schedule.

[0040] Multiplying the calculated base impact coefficient of 0.45, efficiency correction coefficient of 1.2, and time delay correction coefficient of 1.05, we obtain the critical path impact coefficient for Li Si's core tube shear wall reinforcement binding operation: 0.45 multiplied by 1.2 multiplied by 1.05 equals 0.567. This value comprehensively reflects the combined impact of the time sensitivity of the work unit itself, worker efficiency differences, and arrival delays caused by real-time location on the critical path. The larger the value, the more beneficial it is to ensure the project schedule if the worker is assigned to that work unit. Similarly, this coefficient is calculated for each other candidate worker to provide a quantitative basis for subsequent global optimization.

[0041] In one specific implementation, taking the butt welding operation of the steel columns on the seventh floor of a commercial complex project as an example, the calculation process of the process quality coupling degree is explained in detail. Based on the process logic in the Building Information Model, the number of subsequent processes for the current work unit is determined, and the unit rework cost of the corresponding component type for the current work unit is obtained from the preset rework cost library. The product of the number of subsequent processes and the unit rework cost is divided by the preset benchmark value to obtain the process dependence strength coefficient. Specifically, the position of the butt welding operation unit of the steel columns on the seventh floor of the core tube in the process network is extremely critical. Its subsequent processes include ultrasonic flaw detection, anti-corrosion primer application, fireproof coating spraying, installation of external curtain wall connectors, and floor slab laying, totaling five processes. Once a quality defect occurs in this welding operation, all subsequent processes will be suspended or reworked. The unit rework cost for the steel column butt welding is found to be 8,000 yuan per weld, which includes all costs such as carbon arc gouging to remove defects, reheating welding, heat treatment, and flaw detection re-inspection. Multiplying the number of subsequent processes (5) by the unit rework cost of 8,000 yuan yields 40,000 yuan. Dividing this by the preset benchmark value of 20,000 yuan gives a process dependence coefficient of 2.0, indicating that the quality of this work unit has a significant impact on the overall project duration and cost.

[0042] Based on the historical quality traceability data of the corresponding laborers stored in the basic information module, the first-pass yield rate of the laborers' work records with the same component type as the current work unit within a preset time window is used as the skill reliability coefficient. Taking welder Zhang Ming as an example, the historical quality traceability data stored in his labor contract information shows that in the past twelve-month time window, Zhang Ming completed 58 steel column butt welding operations, of which 54 passed ultrasonic flaw detection on the first attempt, and the other four had defects of varying degrees requiring rework. Based on this, his first-pass yield rate is calculated to be 93.1%. Welder Wang Lei, who is a candidate for the same position as Zhang Ming, completed 42 similar operations within the same time window, passing 37 on the first attempt, with a pass rate of 88.1%; welder Li Qiang completed 36 operations, passing 35 on the first attempt, with a pass rate as high as 97.2%. The above first-pass yield rates are directly used as the skill reliability coefficients, i.e., Zhang Ming's is 0.931, Wang Lei's is 0.881, and Li Qiang's is 0.972.

[0043] Based on the real-time location data collected by the real-time sensing module, the distance between the worker's current location and the center point of the work surface corresponding to the current work unit in the Building Information Model (BIM) is calculated. The distance is then converted into a position compliance coefficient within the range [0,1] using a preset distance-compliance mapping function. Taking welder Zhang Ming as an example, the real-time sensing module detects his current location coordinates as a dormitory building in the living area. The coordinates of the center point of the work surface of the seventh-floor steel column in the core tube of the BIM are known. The path planning algorithm calculates the shortest feasible distance between the two points to be 220 meters. The system presets a first threshold of 5 meters and a second threshold of 100 meters for the distance mapping. That is, when the worker is within 5 meters of the work surface, he is considered to be in position, and the position compliance coefficient is 1.0; when the distance exceeds 100 meters, it is considered too far and may affect timely arrival, and the position compliance coefficient is 0.0; when the distance is between 5 meters and 100 meters, the position compliance coefficient decreases linearly with increasing distance. Zhang Ming's current distance of 220 meters exceeds 100 meters, therefore his position compliance coefficient is 0.0. Another welder, Wang Lei, is currently located in the material warehouse, 60 meters from the work surface. His distance falls between 5 and 100 meters. Linear interpolation yields a position accuracy coefficient of (100-60) / (100-5) = 0.42, or 0.42. Welder Li Qiang is currently located on the sixth floor of the core tube work surface, 4 meters vertically from the seventh floor steel column work surface. Including the passageway distance, the total distance is 8 meters. Since 8 meters is greater than 5 meters, his position accuracy coefficient is calculated to be (100-8) / (100-5) = 0.97, or 0.97.

[0044] Based on environmental data collected by the real-time sensing module, the deviations between the current ambient temperature, humidity, and wind speed and the standard values ​​of the process environment requirements corresponding to the current work unit in the Building Information Model (BIM) are calculated. After normalizing each deviation value, the maximum value is taken as the environmental deviation index. Subtracting the environmental deviation index from one yields the environmental compliance coefficient. The BIM specifies that for steel column butt welding operations, the process environment requirements are: ambient temperature not lower than 5 degrees Celsius, ambient humidity not higher than 80%, and ambient wind speed not higher than 2 meters per second. The real-time sensing module collected data showing that the ambient temperature at the current core tube seventh-floor work surface is 2 degrees Celsius, ambient humidity is 75%, and ambient wind speed is 2.5 meters per second. For the temperature parameter, a measured value of 2 degrees Celsius is 5 degrees Celsius below the required lower limit, with a deviation of 3 degrees Celsius. When normalizing, using 10 degrees Celsius as the baseline range, the temperature deviation index is calculated as 3 divided by 10, which equals 0.3. For the humidity parameter, a measured value of 75% is 80% below the upper limit, with a deviation of 5%. When normalizing, using 20% ​​as the baseline range, the humidity deviation index is calculated as 0.05 divided by 0.2, which equals 0.25. For the wind speed parameter, a measured value of 2.5 meters per second exceeds the upper limit of 2 meters per second, with a deviation of 0.5 meters per second. When normalizing, using 3 meters per second as the baseline range, the wind speed deviation index is calculated as 0.5 divided by 3, approximately equal to 0.167. Taking the maximum of the three, 0.3, as the environmental deviation index, and subtracting 0.3 from 1, yields an environmental compliance coefficient of 0.7. If the measured value of a parameter significantly exceeds the limit, for example, if the wind speed reaches 4 meters per second, the wind speed deviation index may exceed 1.0. In this case, the environmental compliance coefficient will be truncated to 0.0.

[0045] Multiplying the calculated process dependency strength coefficient, skill reliability coefficient, position conformity coefficient, and environmental conformity coefficient by the process quality coupling degree yields the process quality coupling degree. Taking welder Li Qiang as an example, his process dependency strength coefficient is 2.0, skill reliability coefficient is 0.972, position conformity coefficient is 0.97, and environmental conformity coefficient is 0.7. Multiplying these four factors gives 2.0 × 0.972 × 0.97 × 0.7, which equals 1.321. Welder Zhang Ming's position conformity coefficient is 0.0, resulting in a final product of zero, indicating that he is currently unable to perform the task. Welder Wang Lei's coefficients are 2.0, 0.881, 0.42, and 0.7, respectively, multiplied to 0.519. The process quality coupling degree comprehensively reflects the impact of the task unit's own quality sensitivity, the worker's historical skill level, real-time availability, and environmental suitability on the final quality. A higher value indicates a greater probability that the worker can complete the task unit and ensure quality under current conditions. Based on this, the system will prioritize assigning highly coupled workers to critical processes with stringent quality requirements in subsequent global optimizations, thereby reducing quality risks at the source.

[0046] In one specific implementation, taking the concrete pouring of the eight-story shear wall of the core tube of a super high-rise building and the hoisting of the outer mega-column steel structure as two conflicting work units as examples, the process of calculating the suitability of candidate workers using fuzzy logic rules is described in detail. The worker state risk index, the critical path influence coefficient, and the process quality coupling degree are used as three input variables of the fuzzy logic system. The precise values ​​of each input variable are converted into corresponding fuzzy values ​​according to a preset membership function. Taking candidate worker Wang Wu as an example, for the core tube concrete pouring work unit, the previous steps calculated his worker state risk index to be 1.28, the critical path influence coefficient to be 1.35, and the process quality coupling degree to be 0.42. The system presets a triangular or trapezoidal membership function for each input variable to map the precise value to a membership degree of low, medium, or high linguistic values. For the worker status risk index, its universe of discourse is set to 0.5 to 2.0, where 0.5 to 1.0 corresponds to low, 0.8 to 1.5 to medium, and 1.2 to 2.0 to high. Wang Wu's 1.28 is calculated to belong to the medium level with a degree of 0.8, to the high level with a degree of 0.2, and to the low level with a degree of 0. For the critical path influence coefficient, its universe of discourse is 0.8 to 2.0, where 0.8 to 1.2 is low, 1.0 to 1.6 is medium, and 1.4 to 2.0 is high. Wang Wu's 1.35 is calculated to belong to the medium level with a degree of 0.75, to the low level with a degree of 0.25, and to the high level with a degree of 0. For the process quality coupling degree, its universe of discourse is 0.2 to 1.5, where 0.2 to 0.7 is low, 0.5 to 1.0 is medium, and 0.8 to 1.5 is high. Wang Wu's 0.42 is calculated to belong to the low level with a degree of 0.86, to the medium level with a degree of 0.14, and to the high level with a degree of 0. This completes the conversion of the three input variables from precise values ​​to fuzzy values.

[0047] The fuzzy values ​​of the three input variables are input into a pre-defined fuzzy rule base for fuzzy inference. This rule base is constructed based on expert experience and historical construction data. Existing technologies typically establish fuzzy rule bases using expert interviews or statistical induction. This specific implementation combines the Delphi method with historical case mining techniques. A panel of five construction project managers with over twenty years of experience was invited to independently evaluate the impact of various combinations of three factors—worker status risk index, critical path influence coefficient, and process quality coupling degree—on the suitability. After four rounds of feedback convergence, initial rules were formed. Simultaneously, five hundred successful work assignment cases were extracted from the company's historical database. Rough set theory was used to mine the association rules between factors and results to verify and supplement the expert rules. The final rule base contains twenty-seven rules. For example, Rule 1: If the worker's state risk index is low, the critical path impact coefficient is high, and the process quality coupling degree is high, then the fit is high; Rule 12: If the worker's state risk index is medium, the critical path impact coefficient is medium, and the process quality coupling degree is medium, then the fit is medium; Rule 18: If the worker's state risk index is high, the critical path impact coefficient is low, and the process quality coupling degree is low, then the fit is low; Rule 21: If the worker's state risk index is medium, the critical path impact coefficient is high, and the process quality coupling degree is low, then the fit is relatively low; Rule 25: If the worker's state risk index is low, the critical path impact coefficient is medium, and the process quality coupling degree is medium, then the fit is relatively high. The fuzzy values ​​of Wang Wu are input into the rule base, triggering several rules, including Rule Twelve and Rule Eighteen. For example, the combination of the preconditions "medium," "medium," and "medium" in Rule Twelve partially matches Wang Wu's case, and its activation strength is the minimum of the membership degrees of the three preconditions, which is 0.75. In the combination of the preconditions "medium," "high," and "low" in Rule Twenty-One, the influence coefficient of the critical path of the task belongs to the "high" level, which is zero, so this rule is not activated. Through fuzzy inference, an output fuzzy set is obtained, which contains the activation strengths corresponding to the five linguistic values: low, lower, medium, higher, and high.

[0048] The fuzzy value of the fit is converted into a precise value within the range [0,1] using a preset defuzzification method, thus obtaining the fit degree of the candidate worker for the current conflicting work unit. Existing defuzzification methods mainly include the maximum membership method, the centroid method, and the weighted average method. This specific implementation adopts the centroid method because it can fully utilize the output information of all activation rules, making the defuzzification result smoother and more continuous. The calculation formula for the centroid method is that the precise output value equals the centroid abscissa of the area enclosed by the membership function curves of each output linguistic value. In specific implementation, the output universe of discourse is discretized into one hundred sampling points. The membership degree at each sampling point is taken as the maximum value output by all activation rules at that point. The sum of the products of the abscissa of each sampling point and the membership degree is calculated and divided by the sum of the membership degrees. Taking Wang Wu as an example, after fuzzy inference, the activation intensity of the output fuzzy set belonging to the middle category is 0.75, and the activation intensity of the higher category is 0.2. The peak value of the membership function belonging to the higher category is at 0.75, and the peak value of the middle category is at 0.5. The exact fit degree calculated by the centroid method is 0.62. This value is between the middle and higher categories, comprehensively reflecting the actual situation that Wang Wu's state risk is moderate, the critical path contribution is high, and the quality coupling degree is low. Similarly, when calculating the fit degree of Wang Wu for the peripheral giant column hoisting operation unit, its worker state risk index is still 1.28, the task critical path influence coefficient is 0.92, and the process quality coupling degree is 0.78. After fuzzification, the task critical path influence coefficient belongs to the lower category to a higher degree, and the process quality coupling degree belongs to the middle category to a higher degree. After inputting into the rule base and triggering different rules, the defuzzification yields a fit degree of 0.58.

[0049] The aforementioned fuzzification, fuzzy inference, and defuzzification processes are performed on each worker in the candidate set. For example, candidate worker Qian Yongjun's three indices for core tube concrete pouring are 1.05, 1.18, and 0.65, respectively, with a calculated fit of 0.71; for peripheral mega-column hoisting, they are 1.05, 0.88, and 0.81, with a fit of 0.75. Candidate worker Sun Jianguo's indices for core tube concrete pouring are 1.52, 1.42, and 0.38, with a fit of 0.33; for peripheral mega-column hoisting, they are 1.52, 0.95, and 0.63, with a fit of 0.41. Through the above calculations, the system constructs a fit matrix with candidate workers as rows and conflicting work units as columns, where the core tube concrete pouring work unit corresponds to Wang Wu 0.62, Qian Yongjun 0.71, and Sun Jianguo 0.33, and the peripheral mega-column hoisting work unit corresponds to Wang Wu 0.58, Qian Yongjun 0.75, and Sun Jianguo 0.41. This matrix provides a quantitative basis for subsequent global combinatorial optimization, ensuring that clear and comparable decision data can be extracted from complex situations involving fuzzy and multi-factor intertwined factors.

[0050] In one specific implementation, taking the concrete pouring of the seven-story shear wall of the core tube and the hoisting of the outer mega-column steel structure of a large commercial complex project as examples, the process of generating the optimal work assignment plan by globally combining and optimizing based on the suitability calculation results is described in detail. A suitability matrix is ​​constructed, with the multiple conflicting work units as rows and the multiple laborers in the candidate personnel set as columns. Each element in this matrix represents the suitability of the corresponding laborer for the corresponding work unit. In this implementation, the core tube concrete pouring work unit is labeled as work unit A, and the outer mega-column hoisting work unit is labeled as work unit B. The candidate personnel set includes five laborers selected in the previous steps: worker A, worker B, worker C, worker D, and worker E. The preliminary steps used fuzzy logic rules to calculate the suitability of each worker for the two work units: Worker A's suitability for work unit A was 0.62, and for work unit B it was 0.58; Worker B's suitability for work unit A was 0.71, and for work unit B it was 0.75; Worker C's suitability for work unit A was 0.33, and for work unit B it was 0.41; Worker D's suitability for work unit A was 0.68, and for work unit B it was 0.52; Worker E's suitability for work unit A was 0.54, and for work unit B it was 0.63. The resulting suitability matrix is ​​a two-dimensional matrix with work units A and B as two rows and five workers as five columns. The matrix elements are the values ​​mentioned above. This matrix quantitatively characterizes the matching degree between each candidate worker and each conflicting work unit, providing basic data for subsequent optimization.

[0051] An integer programming model is constructed with the objective function of maximizing the sum of the fit of all dispatched workers to their assigned work units, and with safety baseline constraints, schedule constraints, and quality requirements as constraints. The objective function is expressed as maximizing the sum of the products of each dispatch decision variable and its corresponding fit. Each worker can be dispatched to at most one work unit. The number of workers required for each work unit is determined based on the workload. In this implementation, work unit A requires two concrete workers, and work unit B requires one crane operator and one signalman. Since the candidate pool includes personnel with the corresponding skills, work unit A requires two workers, and work unit B requires two workers. The safety baseline constraint requires that the worker status risk index of each worker must be lower than a preset risk threshold. This threshold is set based on national occupational health standards and enterprise safety management systems. Existing technologies typically use occupational health risk assessment methods, such as the ergonomic risk index grading standard proposed by the National Institute for Occupational Safety and Health (NIOSH) in the United States. Considering the characteristics of the construction industry, the risk threshold is set at 1.5, meaning that a worker status risk index below 1.5 is considered acceptable risk, while a risk index above 1.5 prohibits dispatching workers to high-altitude or high-intensity positions. The calculations in the previous steps show that worker C's worker status risk index is 1.52, which exceeds the threshold of 1.5. Therefore, the possibility of him being assigned to any work unit must be excluded in the constraints.

[0052] The schedule constraint is that each work unit must be completed within its time window, and the total arrival time of all dispatched workers must not exceed the maximum allowable delay for the planned start time. The time window is derived from the Building Information Model (BIM). The time window for work unit A is from 8:00 AM to 5:00 PM on the same day, and the time window for work unit B is from 9:00 AM to 4:00 PM on the same day. The maximum allowable delay for the planned start time is determined according to the construction organization design. The allowable delay for critical path work units is no more than 15 minutes, and for non-critical path work units, it is no more than 30 minutes. In this implementation, work unit A is on the critical path, so the maximum allowable delay is 15 minutes, and work unit B is on the non-critical path, so the maximum allowable delay is 30 minutes. Arrival time is calculated based on real-time location. For example, if worker A is 300 meters away from work unit A and takes 200 seconds to walk, and 450 meters away from work unit B and takes 300 seconds to walk, the total arrival time of all personnel dispatched to the same work unit must not exceed the maximum allowable delay for that work unit.

[0053] The quality requirement is that the process quality coupling degree of the workers assigned to each work unit must be higher than the preset quality threshold. This threshold is set based on current national construction quality acceptance standards and the company's internal quality control standards. Existing technologies typically employ risk-based quality control methods. Referring to the Project Management Institute's (PMI) quality management knowledge system and combining it with the quality risk levels of each process in the Building Information Model (BIM), the quality threshold for critical processes is set at 0.35, and for general processes at 0.25. In this embodiment, work unit A is the core tube concrete pouring, a critical process, with a quality threshold set at 0.35. Work unit B is the hoisting of the outer mega-columns, also a critical process, with the same quality threshold set at 0.35. Previous calculations show that worker C's process quality coupling degree for work unit A is 0.38, but he has been excluded because his status risk index exceeds the standard. Worker D's process quality coupling degree for work unit A is 0.33, lower than the 0.35 threshold. Therefore, worker D cannot be assigned to work unit A in the work assignment decision, but assignment to work unit B can be considered.

[0054] Solving the integer programming model yields the optimal matching relationship between each conflicting work unit and its assigned workers, as well as a list of unassigned workers. The integer programming model can be solved using operations research methods such as branch and bound or the Hungarian algorithm; this implementation uses a business optimization solver. Input data includes a fitness matrix, the number of workers required for each work unit, and the skills and constraints of each worker. The optimal matching results are as follows: Work unit A assigns workers B and D, with fitness levels of 0.71 and 0.68 respectively, summing to 1.39; Work unit B assigns workers A and E, with fitness levels of 0.58 and 0.63 respectively, summing to 1.21; the total fitness sum is 2.60. Upon inspection, worker B's worker status risk index of 1.05 is below the threshold of 1.5, and worker D's worker status risk index of 1.18 is also below the threshold of 1.5. Their arrival times to work unit A are 240 seconds and 180 seconds respectively, totaling 420 seconds (7 minutes), which does not exceed the allowable delay limit of 15 minutes. Their process quality coupling degrees are 0.65 and 0.33 respectively. Although worker D's 0.33 is slightly below the threshold of 0.35, this solution is still the optimal feasible solution in integer programming due to other constraints. Workers A and E have worker status risk indices of 1.28 and 1.15 respectively, both below the threshold. Their arrival times to work unit B are 300 seconds and 240 seconds respectively, totaling 540 seconds (9 minutes), which does not exceed the allowable delay limit of 30 minutes. Their process quality coupling degrees are 0.78 and 0.63 respectively, both above the threshold of 0.35. Worker C, who was not assigned work, was excluded from the model because his worker status risk index of 1.52 exceeds the safety baseline threshold.

[0055] Based on the solution results, an optimal work assignment plan is generated, including personnel identification, work unit identification, work time, work surface location, and process requirements. This plan is then pushed to the terminal devices of relevant personnel. Taking worker B as an example, their work assignment plan clearly states: Worker B is assigned to the concrete pouring of the shear wall on the seventh floor of the core tube of the work unit. The work time is from 8:00 AM to 5:00 PM on the same day. The work surface location is the coordinate area of ​​the seventh floor of the core tube. The process requirements are: slump control between 180 mm and 220 mm, vibration spacing no greater than 500 mm, no cold joints, and continuous pouring. A message is also pushed to remind them to arrive at the work surface on time and to inform them that their partner is worker D. Worker A's work assignment plan clearly states: Worker A is assigned to the hoisting of the outer mega-column steel structure of the work unit. The work time is from 9:00 AM to 4:00 PM on the same day. The work surface location is the area between mega-columns G12 and G15. The process requirements are: verticality deviation control within one-thousandth, final tightening torque of high-strength bolts reaching the design value, and coordination with signalman worker E. Worker C, who was not assigned any work, received a system notification stating that he was temporarily not scheduled for work because his physiological parameters for the day exceeded the safety threshold, and was advised to rest and observe. Through the above global optimization, the system achieved optimal allocation of human resources under multiple constraints of safety, schedule, and quality, effectively resolving the resource conflict problem among multiple work units.

[0056] In one specific implementation, taking the resource conflict and automatic optimization failure of two work units—the concrete pouring of the eight-story shear wall of the core tube and the hoisting of the outer mega-column steel structure—in a large commercial complex project as an example, the guidance intervention process is described in detail. After completing the global combination optimization, it was found that the safety baseline, schedule constraint, and quality requirements could not be met simultaneously. In this implementation, the candidate personnel set includes five workers, denoted as worker A, worker B, worker C, worker D, and worker E. The core tube concrete pouring work unit is labeled as work unit A, and the outer mega-column hoisting work unit is labeled as work unit B. The previous steps have completed the construction of the fit degree matrix and the solution of the integer programming model, but the solution results show that all possible work assignment combinations cannot simultaneously satisfy the three constraints. Specifically, if worker A and worker B are assigned to work unit A, worker A's worker status risk index is 1.52, exceeding the preset risk threshold of 1.5, thus violating the safety baseline constraint. If worker C and worker D are assigned to work unit A, their combined arrival time at the work site is 22 minutes, exceeding the allowable delay limit of 15 minutes for work unit A, thus violating the schedule constraint. If worker B and worker E are assigned to work unit A, worker E's process quality coupling degree for work unit A is 0.32, lower than the preset quality threshold of 0.35, thus violating the quality requirement. Other combinations also present similar problems, and work unit B also cannot find a work assignment scheme that satisfies all constraints. After traversing calculations, the system confirms that under the current candidate personnel set and constraints, there is no feasible solution, thus triggering the guidance intervention process.

[0057] All relevant information generated during the task decomposition, intelligent matching, and dynamic control phases is summarized and structured to form a complete guidance intervention data package. This data package includes detailed task information for work unit A and work unit B generated during the task decomposition phase. Work unit A requires two senior concrete workers, with a time window from 8:00 AM to 5:00 PM on the same day, and quality requirements of a slump control between 180 mm and 220 mm with no cold joints. Work unit B requires one intermediate crane operator and one intermediate signalman, with a time window from 9:00 AM to 4:00 PM on the same day, and quality requirements of verticality deviation within 0.1% and the final tightening torque of high-strength bolts reaching the design value. The intelligent matching phase generates a set of candidate personnel information, including the skill qualifications, real-time location, and current availability of five workers. Workers A and C hold senior concrete worker certificates, workers B and D hold intermediate crane operator certificates and are also qualified signalmen, and worker E holds an intermediate signalman certificate. The core indicator data for each worker in each work unit, calculated during the dynamic control phase, includes worker status risk index, critical path impact coefficient, process quality coupling degree, and final fit degree. For example, worker A's status risk index is 1.52, process quality coupling degree for work unit A is 0.48, and fit degree is 0.62; worker C's status risk index is 1.18, process quality coupling degree for work unit A is 0.65, and fit degree is 0.71; worker E's status risk index is 1.15, process quality coupling degree for work unit B is 0.63, and fit degree is 0.63, etc. The data package also includes records of constraint violations during the optimization model's solution process. For example, safety baseline violation records show worker A's status risk index exceeding the limit; schedule violation records show the arrival time of the combination of workers C and D exceeding the allowable upper limit; and quality violation records show worker E's process quality coupling degree for work unit A not meeting the threshold requirement, etc.

[0058] The aforementioned guidance and intervention data package is pushed to the mobile terminal device held by the project leader, and key information is presented in a visual manner to facilitate the leader's quick understanding of the essence of the problem. The push interface uses a 3D digital twin scene as the base map, highlighting and flashing the work unit A at the eighth-floor work surface of the core tube, and highlighting and flashing the work unit B in the outer giant pillars G12 to G15 area. At the same time, it displays the real-time location icons and key status labels of five candidate personnel in a list format. For example, a red warning of excessive risk is marked next to worker A's head, a yellow warning of distance is marked next to worker C's head, and a yellow warning of quality risk is marked next to worker E's head. The left side of the interface displays constraint violation details in layers. The safety baseline layer shows that worker A's status risk index of 1.52 exceeds the threshold of 1.5. The schedule layer shows that the arrival time of worker C and worker D combined is 22 minutes, exceeding the 15-minute limit. The quality layer shows that worker E's process quality coupling degree for work unit A is 0.32, which is lower than the threshold of 0.35. The bottom of the interface provides a variety of hypothesis analysis tools. The leader can try to adjust different constraint thresholds or recombine personnel. The system calculates and provides feedback on the constraint satisfaction status in real time. For example, if the person in charge tries to temporarily exclude worker A, the system automatically recalculates the remaining personnel combinations and finds that worker C and worker D still have a problem with exceeding the time limit; if the person in charge tries to extend the time window of work unit A by one hour, the system calculation shows that the arrival time of worker C and worker D becomes eighteen minutes, which still exceeds the fifteen-minute limit; if the person in charge tries to increase the allowed delay limit to twenty minutes, the system prompts that worker C and worker D can meet the time requirement at this time, but it is necessary to confirm whether to accept the impact on subsequent processes.

[0059] The project manager makes the final decision based on the comprehensive information provided by the system and their own on-site experience, and issues guidance instructions through the terminal device. In this implementation, the manager makes the following decisions after comprehensively considering the actual situation on site: First, worker A is temporarily transferred from the core tube area to perform ground material handling work, as their condition risk index exceeds the standard and they are not suitable for high-altitude work; Second, a worker with an advanced concrete worker certificate is urgently transferred from an adjacent section and added to the candidate pool. The worker's real-time location shows that they are only 80 meters away from the core tube working face, with a condition risk index of 0.95 and a process quality coupling degree of 0.72; Third, the planned start time of work unit A is adjusted from 8:00 AM to 8:30 AM to provide workers C and D with more time to arrive. After the person in charge completes the above adjustments on the terminal interface, the system immediately performs integer programming again with the updated data, obtaining a new feasible work assignment scheme: Work unit A is assigned to workers F and C, whose status risk indices are 0.95 and 1.18 respectively, both below the threshold. The sum of their arrival times is 40 seconds for 80 meters plus 200 seconds for 300 meters, totaling 240 seconds or 4 minutes, which is far below the upper limit of the delay corresponding to the adjusted time window. Their process quality coupling degrees are 0.72 and 0.65 respectively, both above the threshold of 0.35. Work unit B is assigned to workers B and E, whose status risk indices are 1.05 and 1.15 respectively, both below the threshold. The sum of their arrival times is 100 seconds for 150 meters plus 130 seconds for 200 meters, totaling 230 seconds or 3.8 minutes, which does not exceed the upper limit of 30 minutes. Their process quality coupling degrees are 0.78 and 0.63 respectively, both above the threshold. After the person in charge confirms the scheme, they click "Issue," and the system will immediately push the formal work assignment instructions to the terminal devices of workers F, C, B, and E, while notifying worker A of the temporary adjustment arrangements. Through the aforementioned guidance and intervention process, in extreme cases where the algorithm cannot solve the problem automatically, the system leverages the experience and flexible decision-making of human experts to effectively handle complex on-site issues, ensuring the coordinated unity of project safety, schedule, and quality.

[0060] The above-mentioned models or function formulas are all dimensionless and numerical calculations. The models or function formulas are obtained by software simulation based on a large amount of collected data to obtain the most recent real situation. The preset parameters in the models or function formulas are set by those skilled in the art according to the actual situation.

[0061] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0062] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0063] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0064] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A labor dispatch system for construction labor subcontracting, characterized in that, include: The basic information module is used to store and manage the identity information, skill qualifications, and labor contract information of the workers. The real-time sensing module is used to collect the real-time location, physiological parameters, and environmental data of the workers' work site through wearable terminal devices. The task decomposition module is used to decompose engineering tasks into work units that include corresponding skill requirements, time windows, and quality requirements based on the construction schedule imported from the preset building information model. The intelligent matching module is used to perform preliminary matching based on the skill requirements of the work unit and the skill qualifications of the workers, and to filter out the workers who are in an idle state by combining real-time location to form a candidate set. The dynamic control module is used to calculate the worker status risk index based on the physiological status parameters collected by the real-time perception module when the candidate personnel set output by the intelligent matching module cannot simultaneously meet the priority requirements of multiple work units and there are resource conflicts. Based on the schedule plan and process logic in the building information model, it calculates the critical path influence coefficient of the task and the process quality coupling degree. Taking the worker status risk index, the critical path influence coefficient of the task and the process quality coupling degree as input parameters, it calculates the suitability of each candidate worker for the conflicting work units through fuzzy logic rules, and performs global combination optimization based on the suitability calculation results to generate the optimal work assignment plan that meets the safety bottom line, schedule constraints and quality requirements. The guidance and intervention module is used to push information from the task decomposition module, intelligent matching module, and dynamic control module to the project leader when safety bottom lines, schedule constraints, and quality requirements cannot be met, serving as the basis for their actual guidance and intervention.

2. A labor dispatch system for construction labor subcontracting according to claim 1, characterized in that, The identity information includes the worker's historical health records and age data; the skills and qualifications include the worker's skills and corresponding levels; and the labor contract information includes the worker's historical performance data and historical quality traceability data.

3. A labor dispatch system for construction labor subcontracting according to claim 2, characterized in that, The pre-defined building information model is a structured database containing three-dimensional component geometric data, process logic relationships, planned start time, planned completion time, and pre-defined quality acceptance standards.

4. A labor dispatch system for construction labor subcontracting according to claim 3, characterized in that, Decomposing engineering tasks into work units that include corresponding skill requirements, time windows, and quality requirements refers to: Based on the process logic in the building information model, the unit project is divided into several construction sections. The required labor skills and corresponding levels are matched from the preset skill library according to the component type corresponding to the construction section to form skill requirements. The workable time interval is determined by combining the planned start time and planned completion time with the process logic to form a time window. The corresponding process requirements and acceptance indicators are extracted from the preset quality library according to the preset quality acceptance standards to form quality requirements.

5. A labor dispatch system for construction labor subcontracting according to claim 4, characterized in that, In the dynamic control module, the worker status risk index is calculated by combining the physiological status parameters collected by the real-time sensing module and the environmental data of the work site with the historical health records and age data of the corresponding workers stored in the basic information module. The critical path impact coefficient is calculated based on the schedule in the building information model, combined with the historical performance data contained in the labor contract information of the corresponding laborers stored in the basic information module, and the time taken to reach the work surface reflected by the real-time location collected by the real-time sensing module. The process quality coupling degree is obtained by correlation analysis based on the process logic in the building information model, combined with the historical quality traceability data contained in the labor contract information of the corresponding laborers stored in the basic information module, and the degree of conformity of the work surface and the degree of environmental suitability reflected by the real-time location and environmental data collected by the real-time sensing module.

6. A labor dispatch system for construction labor subcontracting according to claim 5, characterized in that, The worker condition risk index is calculated using the following steps: Based on the parameters collected by the real-time sensing module, according to the preset benchmark interval and safety threshold, the measured values ​​are mapped to single risk factors in the [0,1] interval through a piecewise linear function: if the parameter has a benchmark interval, the risk factor is 0 when the measured value is within the benchmark interval, the risk factor is proportional to the degree of exceeding the interval but not exceeding the safety threshold when it exceeds the safety threshold; and it is always 1 when it exceeds the safety threshold. If the parameter only has a safety threshold, the risk factor is proportional to the measured value when the measured value is below the safety threshold, and is always 1 when the measured value exceeds the safety threshold. The physiological load index is obtained by weighting and summing the risk factors of heart rate and body temperature; the environmental stress index is obtained by weighting and summing the risk factors of temperature, humidity, and dust. Based on the age stored in the basic information module, the age correction coefficient is obtained according to the preset age-risk coefficient mapping table; Based on the historical health records stored in the basic information module, the health correction coefficient is determined according to the number of basic disease types recorded therein. The physiological stress index and the environmental stress index are added together, and then multiplied by the product of the age correction factor and the health correction factor to obtain the worker's condition risk index.

7. A labor dispatch system for construction labor subcontracting according to claim 5, characterized in that, The critical path impact coefficient of the task is calculated according to the following steps: Based on the schedule in the building information model, the planned start time and planned completion time of the current work unit are extracted, the total float time of the work unit is calculated, and the foundation influence coefficient is determined linearly according to the length of the total float time. Based on the historical performance data contained in the labor contract information of the corresponding laborers stored in the basic information module, the ratio of the laborer's average work efficiency to the standard work efficiency is calculated and used as the efficiency correction coefficient. Based on the real-time location collected by the real-time sensing module, the path distance between the worker's current location and the work surface of the conflicting work unit is calculated. The arrival time is estimated according to the preset moving speed, and the time delay correction coefficient is determined based on the difference between the arrival time and the planned start time. Multiply the base impact coefficient by the efficiency correction coefficient, and then multiply by the time delay correction coefficient to obtain the critical path impact coefficient of the mission.

8. A labor dispatch system for construction labor subcontracting according to claim 5, characterized in that, The process quality coupling degree is calculated according to the following steps: Based on the process logic in the building information model, the number of subsequent processes for the current work unit is determined, and the unit rework cost of the corresponding component type for the current work unit is obtained from the preset rework cost library. The product of the number of subsequent processes and the unit rework cost is divided by the preset benchmark value to obtain the process dependence strength coefficient. Based on the historical quality traceability data of the corresponding laborers stored in the basic information module, the pass rate of the laborer's work records with the same component type as the current work unit is statistically analyzed within a preset time window, and used as the skill reliability coefficient. Based on the real-time location collected by the real-time sensing module, the distance between the current location of the worker and the center point of the work surface corresponding to the current work unit in the building information model is calculated. The distance is converted into a location conformity coefficient in the range of [0,1] according to the preset distance-conformity mapping function. The distance-conformity mapping function is as follows: when the distance is less than or equal to the first threshold, the location conformity coefficient is 1; when the distance is greater than or equal to the second threshold, the location conformity coefficient is 0; when the distance is between the first threshold and the second threshold, the location conformity coefficient decreases linearly with the increase of distance. Based on the environmental data collected by the real-time sensing module, the deviation values ​​between the current ambient temperature, ambient humidity, and ambient wind speed and the standard values ​​of the process environment requirements corresponding to the current work unit in the building information model are calculated respectively. After normalizing each deviation value, the maximum value is taken as the environmental deviation index. The environmental compliance coefficient is obtained by subtracting the environmental deviation index from 1. The environmental compliance coefficient is truncated to the interval [0,1]. The process quality coupling degree is obtained by multiplying the process dependence intensity coefficient, skill reliability coefficient, location conformity coefficient, and environment conformity coefficient.

9. A labor dispatch system for construction labor subcontracting according to claim 8, characterized in that, The suitability of each candidate worker for the conflicting work unit is calculated using fuzzy logic rules, specifically as follows: The worker status risk index, the critical path influence coefficient of the task, and the process quality coupling degree are used as the three input variables of the fuzzy logic system. The precise values ​​of each input variable are converted into corresponding fuzzy values ​​according to the preset membership function. The fuzzy values ​​include three linguistic values: low, medium, and high. The fuzzy values ​​of the three input variables are input into a preset fuzzy rule base for fuzzy inference. The inference rules output corresponding fuzzy values ​​of fit degree based on different combinations of worker state risk index, task critical path influence coefficient and process quality coupling degree. The fuzzy values ​​of fit degree include five linguistic values: low, lower, medium, higher and high. The fuzzy value of the fit is converted into a precise value in the range [0,1] using a preset defuzzification method, so as to obtain the fit of the candidate worker with respect to the current conflicting work unit.

10. A labor dispatching system for construction labor subcontracting according to claim 9, characterized in that, Based on the results of the fit calculation, a global combination optimization is performed to generate an optimal work assignment plan that meets the safety baseline, schedule constraints, and quality requirements. Specifically: Construct a fit degree matrix with multiple work units currently in conflict as rows and multiple laborers in the candidate set as columns. Each element in the fit degree matrix represents the fit degree of the corresponding laborer for the corresponding work unit. An integer programming model is constructed with the objective function of maximizing the sum of the fit of all dispatched workers to their assigned work units, and with safety baseline constraints, schedule constraints, and quality requirements as constraints. Solve the integer programming model to obtain the optimal matching relationship between each conflicting work unit and the assigned workers, as well as the list of unassigned workers.

Images