Partial project closed-loop operation management system and method based on Internet of Things

Abstract

The invention relates to the field of project management, and discloses a system and a method for managing closed-loop operation of a split item project based on the Internet of Things, and the system comprises a task classification module which is used for classifying operation tasks of the split item project; the fitness evaluation module is used for calculating the fitness of the members of the construction team in the aspects of criticality, timeliness and authority of each type of operation tasks, and carrying out fitness fusion to obtain comprehensive evaluation fitness; the member distribution module is used for performing cooperative analysis on the fitness and the social behavior characteristics of the construction team members based on a graph neural network, and optimizing the distribution result of the construction team members; and the closed-loop management module is used for dynamically adjusting the engineering operation mode according to the progress condition of the engineering operation task in combination with the real-time environment data of the engineering site and the distribution result of the construction team personnel. According to the invention, precise matching of engineering operation tasks, optimization of team cooperation and dynamic adaptation of a construction environment are realized.

Description

Technical Field

[0001] The present invention relates to the field of engineering management, and in particular to an Internet of Things-based closed-loop operation management system and method for sub-projects. Background Art

[0002] A construction project management method uses a method to divide the entire project into multiple independent parts and projects based on their functions or work content, facilitating better construction management and control. It breaks down a large project into several smaller parts, each with its own independent construction objectives, requirements, and schedule.

[0003] Engineering Closed-Loop Operation Management (ECOM) ensures that all operational activities are successfully completed within predetermined goals, requirements, and standards through comprehensive, comprehensive monitoring and management of engineering projects. The core of this closed-loop system is continuous feedback and adjustment, ensuring continuous optimization from design to construction and maintenance, thus forming an effective closed-loop management system.

[0004] However, the existing engineering closed-loop operation management system still has the following shortcomings:

[0005] (1) Due to the complexity of construction tasks and the diversity of skills and experience of construction team members, it is difficult to achieve accurate matching in the task allocation process, which may lead to insufficient or excessive allocation of human resources, thereby affecting the quality and efficiency of task execution.

[0006] (2) The efficiency of collaboration between team members directly affects the smoothness of task execution. Under the traditional management model, the collaboration between team members may suffer from poor communication or poor coordination, resulting in a decrease in overall work efficiency.

[0007] (3) How to accurately dispatch personnel based on the timeliness of the work tasks and the members' ability to adjust their time so that the construction project can be completed on time is a complex scheduling problem.

[0008] Currently, no effective solutions have been proposed for the problems in related technologies. Summary of the Invention

[0009] In view of this, the purpose of this application is to provide a closed-loop operation management system and method for sub-projects based on the Internet of Things to overcome the above-mentioned technical problems existing in the existing related technologies.

[0010] To this end, the specific technical solutions adopted in the present invention are as follows:

[0011] According to one aspect of the present invention, a closed-loop operation management system for sub-projects based on the Internet of Things is provided, comprising:

[0012] The task classification module is used to classify the work tasks of each part of the project based on clustering algorithms and work task data collected by the Internet of Things;

[0013] The fitness assessment module is used to calculate the fitness of construction team members in terms of the criticality, timeliness, and authority of each type of work task, and integrate the fitness to obtain a comprehensive fitness assessment;

[0014] The member allocation module is used to collaboratively analyze the comprehensive evaluation fitness, criticality fitness, timeliness fitness, authority fitness, and social behavior characteristics of construction team members based on graph neural networks, and optimize the allocation results of construction team members;

[0015] The closed-loop management module is used to dynamically adjust the engineering operation mode according to the progress of the engineering operation tasks, combined with the real-time environmental data of the engineering site and the allocation results of the construction team personnel.

[0016] Furthermore, based on the clustering algorithm and the task data collected by the Internet of Things, the task classification of each sub-project includes:

[0017] Plan the installation location of IoT monitoring equipment based on the requirements of the sub-projects at the project construction site;

[0018] Using IoT monitoring equipment to collect task-related data and extract task features from the task-related data, where the task-related data includes basic task information, task progress information, task feature data, and task equipment information;

[0019] Based on the clustering algorithm and according to the characteristics of the job tasks, the job tasks are grouped, and the silhouette coefficient is used to evaluate the clustering effect.

[0020] Furthermore, the construction team members' adaptability to the criticality, timeliness, and authority of each type of task is calculated and integrated to obtain a comprehensive evaluation of their adaptability, including:

[0021] Based on the weighted average method and combined with the construction team members' task realization capabilities, the construction team members' adaptability to the critical aspects of each type of work task is calculated;

[0022] Based on the linear regression model and combined with the time adjustment ability of construction team members, the adaptability of construction team members to the timeliness of each type of task is calculated;

[0023] By using fuzzy logic principles and combining the professional capabilities of construction team members, we can quantify the adaptability of construction team members to the authority of each type of work task.

[0024] Based on the construction team members' adaptability to the criticality, timeliness, and authority of each type of task, the comprehensive evaluation adaptability is calculated. The formula for the comprehensive evaluation adaptability is:

[0025]

[0026] Where G ij represents the comprehensive evaluation fitness of construction team member i on task j;

[0027] a1, a2, and a3 represent the importance coefficients of criticality, timeliness, and authority fitness in the comprehensive evaluation of fitness, respectively;

[0028] δ1 and δ2 are both adjustment coefficients;

[0029] Key ij represents the fitness of construction team member i in terms of the criticality of task j;

[0030] Time ij represents the timeliness of construction team member i in task j;

[0031] Au ij represents the adaptability of construction team member i in terms of authority on task j;

[0032] T j A represents the timeliness score of task j, ij represents the time adjustment ability of construction team member i on task j, L i represents the project leadership ability score of construction team member i.

[0033] Furthermore, based on the weighted average method and combined with the construction team members' task realization capabilities, the construction team members' adaptability to the key aspects of each type of task is calculated, including:

[0034] Determine the criticality score of each type of task based on its impact on the completion of each part of the project;

[0035] Obtain the professional capabilities of construction team members associated with each type of work task and their experience matching with each type of work task;

[0036] Calculate the criticality score of each construction team member for each type of task using a weighted scoring method, combined with the professional ability score and experience matching of the construction team members;

[0037] The construction team members' criticality scores for each type of work task are divided by the criticality scores of the corresponding type of work tasks to obtain the construction team members' adaptability to the criticality of each type of work task.

[0038] Furthermore, based on the linear regression model and combined with the time adjustment ability of construction team members, the adaptability of construction team members to the timeliness of each type of task is calculated, including:

[0039] Determine the timeliness score of each type of task based on the execution time of each type of task;

[0040] Obtain the historical completion time and time management ability of construction team members for each type of task, and comprehensively obtain the time adjustment ability of construction team members for each type of task;

[0041] Normalize the timeliness scores of each type of task and standardize the time adjustment ability;

[0042] Construct a linear regression model of time-effectiveness fitness:

[0043]

[0044] Where, represents the adaptability of construction team member i to the timeliness of task category j;

[0045] β0 represents the intercept term, β1 and β2 represent the regression coefficients of the timeliness score of the task and the time adjustment ability of the construction team members, respectively;

[0046] T′ j represents the normalized timeliness score of task category j;

[0047] C′ i represents the time adjustment ability of construction team member i after standardization;

[0048] The linear regression model of timeliness fitness is fitted using the least squares method and historical data, and the target vector is the historical completion time of each type of task completed by the construction team members;

[0049] After obtaining the regression coefficient, it is substituted into the linear regression model of timeliness fitness, and the timeliness fitness of construction team members for each type of work task is calculated through the linear regression model of timeliness fitness.

[0050] Furthermore, by using fuzzy logic principles and combining the professional capabilities of construction team members, we quantify the adaptability of construction team members to the authority of each type of task, including:

[0051] Determine the authority score for each type of task based on its complexity and technical difficulty;

[0052] Obtain the academic background and project leadership ability of construction team members, and use the academic background and project leadership ability as input variables for fuzzy logic reasoning;

[0053] The input variables are converted into fuzzy membership values through membership functions. Based on the fuzzy membership values of academic background and project leadership ability, and through all appropriate fuzzy inference rules and weighted calculation, the fuzzy inference results of construction team members for each type of work task are obtained.

[0054] The defuzzification method is used to convert the fuzzy reasoning results into the authority scores of construction team members for each type of work task;

[0055] The authority score of the construction team members for each type of work task is divided by the authority score of the corresponding type of work task to obtain the adaptability of the construction team members to the authority of each type of work task.

[0056] Furthermore, based on graph neural networks, a collaborative analysis is conducted on comprehensive assessment fitness, criticality fitness, timeliness fitness, authority fitness, and the social behavior characteristics of construction team members. The results of optimizing the allocation of construction team members include:

[0057] Obtaining the social behavior characteristics of construction team members, including the collaboration efficiency among construction team members;

[0058] Construction team members are regarded as nodes, and the node characteristics are fitness in terms of criticality, timeliness, authority, and comprehensive evaluation. The edges and edge weights between nodes are determined according to the social behavior characteristics of construction team members.

[0059] Construct a graph neural network structure and use a message passing mechanism to update node features. The node output of the graph neural network is the final feature vector of each construction team member. The final feature vector includes the construction team member's fitness for a certain type of construction task and social behavior characteristic information after being processed by the graph neural network. The fitness includes fitness in terms of criticality, timeliness, authority, and comprehensive evaluation.

[0060] Based on the gap between the fitness of construction team members for a certain type of construction task and the corresponding construction task requirements, as well as the collaboration efficiency among construction team members, a loss function is constructed. The optimization goal is to allocate work tasks in a way that maximizes the fitness score and improves the collaboration efficiency of the construction team. Gradient descent is used to minimize the loss function, resulting in a trained graph neural network model.

[0061] Through the trained graph neural network model, the job task requirements are input and the final fitness score of each construction team member is output; the construction team members are assigned according to the final fitness score.

[0062] Furthermore, the formula of the loss function is:

[0063]

[0064] Where L represents the total value of the loss function;

[0065] represents the fitness score of construction team member i for task j at the kth iteration, and Ta(j) represents the target requirement of task j;

[0066] w ij represents the weighting coefficient of task j to construction team member i;

[0067] λ represents the weight factor of collaboration efficiency;

[0068] C(S it ) represents the collaboration efficiency between construction team member i and construction team member t;

[0069] n represents the number of construction team members, and m represents the number of work tasks.

[0070] Furthermore, based on the progress of the engineering tasks, combined with the real-time environmental data of the engineering site and the allocation of construction team personnel, the engineering operation mode is dynamically adjusted, including:

[0071] Adjust the allocation of construction team personnel according to the progress of engineering tasks;

[0072] Adjust the allocation of engineering resources based on real-time environmental data at the engineering site and the progress of engineering tasks;

[0073] Obtain the effect of the adjustment of the engineering operation mode and compare the differences before and after the adjustment until the expected engineering operation progress target is achieved.

[0074] According to another aspect of the present invention, there is also provided a closed-loop operation management method for sub-projects based on the Internet of Things, comprising:

[0075] Classify the work tasks of each sub-project based on clustering algorithms and work task data collected by the Internet of Things;

[0076] Calculate the construction team members' adaptability to the criticality, timeliness, and authority of each type of task, and integrate the adaptability to obtain a comprehensive assessment of adaptability;

[0077] Based on graph neural networks, we conduct collaborative analysis of comprehensive assessment fitness, criticality fitness, timeliness fitness, authority fitness, and the social behavior characteristics of construction team members, and optimize the allocation of construction team members.

[0078] Dynamically adjust the engineering operation mode according to the progress of the engineering operation tasks, combined with the real-time environmental data of the engineering site and the allocation results of the construction team personnel.

[0079] The present invention includes the following beneficial effects:

[0080] (1) The present invention uses Internet of Things technology to collect task data in real time, combines clustering algorithms to classify tasks, and evaluates the classification effect, so that the task can be divided into several categories. At the same time, through the weighted average method, linear regression model and fuzzy logic principle, combined with the professional ability, time adjustment ability and experience of the construction team members, the adaptability of the members to the task is quantitatively evaluated. By comprehensively evaluating the construction team members from multiple dimensions (criticality, timeliness and authority), the task allocation is made more accurate, and the advantages of each member can be maximized, thereby improving the overall work efficiency and task completion of the team.

[0081] (2) Graph neural networks are used to collaboratively analyze the fitness and social behavior characteristics of construction team members to optimize task allocation. Graph neural networks optimize collaboration and task allocation among members through a message passing mechanism, significantly improving the efficiency of construction teams when working together. By optimizing the allocation of construction teams, task conflicts and resource waste are reduced, and the overall execution efficiency of the team is improved. By optimizing the loss function, task allocation is ensured to maximize efficiency while improving the overall quality of team collaboration.

[0082] (3) Based on real-time feedback from multiple dimensions, including task progress, environmental data, personnel allocation, and resource allocation, construction operations are dynamically adjusted and continuously optimized through real-time data and analysis results. This achieves a closed-loop management mechanism with real-time feedback and continuous adjustment, ensuring that engineering operations can respond to various changes in a timely manner during the actual process, avoiding delays or waste of resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0084] Figure 1This is a principle block diagram of a closed-loop operation management system for sub-projects based on the Internet of Things according to an embodiment of the present invention;

[0085] Figure 2 The present invention is a flowchart of a closed-loop operation management method for sub-projects based on the Internet of Things.

[0086] In the picture:

[0087] 1. Task classification module; 2. Fitness evaluation module; 3. Member allocation module; 4. Closed-loop management module. DETAILED DESCRIPTION

[0088] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0089] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application pertains. The terms used herein are for the purpose of describing the embodiments of this application only and are not intended to limit this application.

[0090] According to an embodiment of the present invention, a closed-loop operation management system and method for sub-projects based on the Internet of Things are provided.

[0091] The present invention will now be further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to one embodiment of the present invention, a closed-loop operation management system for sub-projects based on the Internet of Things is provided, including:

[0092] The task classification module 1 is used to classify the work tasks of the sub-projects based on the clustering algorithm and the work task data collected by the Internet of Things.

[0093] In one embodiment, based on a clustering algorithm and task data collected by the Internet of Things, the classification of task data for each sub-project includes:

[0094] Plan the installation location of the Internet of Things monitoring equipment based on the requirements of the sub-projects at the project construction site.

[0095] IoT monitoring equipment is used to collect relevant data of the work tasks, and the work task features are extracted from the relevant data of the work tasks. The relevant data of the work tasks include basic information of the work tasks, progress information of the work tasks, task feature data and work equipment information. Specifically, it includes: temperature and humidity, air pressure, air quality, noise, etc. Environmental data is collected to monitor the impact of work conditions on task progress, especially those tasks that have a significant impact on timeliness and safety (such as concrete pouring requires certain temperature and humidity conditions). Task progress information, such as task start time, expected completion time, actual completion time, feedback during task execution, etc. Equipment operating status, equipment failure information, usage time, etc. Sensors can monitor the status of the equipment in real time to ensure that the equipment completes the task within the appropriate time period.

[0096] Based on the clustering algorithm and according to the characteristics of the task, the tasks are grouped. The silhouette coefficient is used to evaluate the clustering effect to ensure that each task group has high internal cohesion and large external separation. The range of the silhouette coefficient is [-1, 1]. The closer the value is to 1, the better the clustering result is; the closer the value is to -1, the poor clustering effect is. Choose an appropriate clustering algorithm, such as K-means, DBSCAN or hierarchical clustering, to group the tasks.

[0097] Types of monitoring equipment include:

[0098] Temperature and humidity sensors: For tasks that require specific environmental conditions (such as concrete pouring, steel structure installation, etc.), temperature and humidity sensors are installed to monitor the construction environment. RFID tags and sensors are installed in material storage areas to monitor the location, quantity, consumption, and other information of materials in real time. Video surveillance cameras: Video surveillance equipment is installed in key areas of the construction site (such as electrical installation areas, high-altitude work areas, etc.) to ensure safety and progress monitoring. Accelerometers and tilt sensors are installed in certain high-risk areas (such as high-altitude work and heavy equipment operation areas) to monitor the stability of equipment and the safety of personnel. The location of equipment should be reasonably planned according to the execution location of each task. For example, steel structure installation and electrical equipment installation tasks are usually performed at high altitudes, and relevant monitoring equipment needs to be installed.

[0099] Material storage area: RFID tags are used to track material consumption and storage status in real time to ensure timely and effective material management.

[0100] The fitness evaluation module 2 is used to calculate the fitness of construction team members in terms of the criticality, timeliness and authority of each type of work task, and to integrate the fitness to obtain a comprehensive evaluation fitness.

[0101] In one embodiment, the construction team members' adaptability to the criticality, timeliness, and authority of each type of task is calculated and integrated to obtain a comprehensive assessment of their adaptability, including:

[0102] Using a weighted average method and combining their task-achievement capabilities, we calculated construction team members' adaptability to the criticality of each task type. Using a linear regression model and combining their time-adjustment abilities, we calculated their adaptability to the timeliness of each task type. Using fuzzy logic principles and combining their professional expertise, we quantified their adaptability to the authority of each task type.

[0103] Based on the construction team members' adaptability to the criticality, timeliness, and authority of each type of task, the comprehensive evaluation adaptability is calculated. The formula for the comprehensive evaluation adaptability is:

[0104]

[0105] Where G ij represents the comprehensive evaluation fitness of construction team member i on task j; a1, a2 and a3 respectively represent the importance coefficients of criticality, timeliness and authority fitness in the comprehensive evaluation fitness; δ1 and δ2 are adjustment coefficients; Key ij represents the critical adaptability of construction team member i in task j; Time ij represents the timeliness adaptability of construction team member i on task j; Au ij T represents the adaptability of construction team member i in terms of authority in task j; j A represents the timeliness score of task j, ij represents the time adjustment ability of construction team member i on task j, L i represents the project leadership ability score of construction team member i.

[0106] In one embodiment, based on the weighted average method and in combination with the task implementation capabilities of the construction team members, the calculation of the construction team members' adaptability to the critical aspects of each type of work task includes:

[0107] Based on the impact of each type of work task on the completion of the sub-item project, determine the criticality score of the corresponding type of work task.

[0108] Obtain the professional capabilities of construction team members associated with each type of work task and their experience matching with each type of work task.

[0109] The criticality scores of construction team members for each type of work task are calculated using a weighted scoring method and combined with the professional ability scores and experience matching of construction team members.

[0110] The construction team members' criticality scores for each type of work task are divided by the criticality scores of the corresponding type of work tasks to obtain the construction team members' adaptability to the criticality of each type of work task.

[0111] The criticality score of each task is determined based on its impact on the overall progress and completion of the project. Generally, tasks that affect project completion are assigned higher criticality scores. Influencing factors include the scale, complexity, resource requirements, time nodes, etc. of the task. For example, infrastructure construction tasks (such as foundation pouring and structural steel frame erection) usually have higher criticality scores, while decoration tasks (such as wall painting) may have lower criticality scores.

[0112] The professional competency score is assigned to each construction team member based on factors such as their individual skills and work experience. For example, Worker A, who has extensive experience in foundation construction, would have a high competency score (e.g., 8 / 10) for this type of task, but a low score (e.g., 5 / 10) for electrical installation. Experience fit is assessed based on the degree to which a construction team member's previous experience matches the requirements of each task, assessing each worker's experience fit for different tasks.

[0113] Each worker's fitness for each type of task is calculated by dividing the construction team member's criticality fitness score by the task's criticality score. The fitness value reflects a member's adaptability to the criticality of a task. A worker with a high adaptability to the criticality requirements of a task (i.e., a high degree of professional competence and experience matching) will have a high fitness value, while a worker with a low fitness value will have a low fitness value.

[0114] In one embodiment, based on a linear regression model and in combination with the time adjustment ability of the construction team members, the construction team members' adaptability to the timeliness of each type of task is calculated, including:

[0115] According to the execution time of each type of task, the timeliness score of the corresponding type of task is determined.

[0116] Obtain the historical completion time and time management ability of construction team members for each type of work task, and comprehensively obtain the time adjustment ability of construction team members for each type of work task.

[0117] The timeliness scores of each type of homework task are normalized and the time adjustment ability is standardized.

[0118] Construct a linear regression model of time-effectiveness fitness:

[0119]

[0120] Where, represents the adaptability of construction team member i to the timeliness of task category j; β0 represents the intercept term, β1 and β2 represent the regression coefficients of the task timeliness score and the construction team member's time adjustment ability, respectively, and are the effects of the task timeliness score and the member's time adjustment ability on the timeliness adaptability score; T′ j represents the normalized timeliness score of task category j; C′ i It represents the time adjustment ability of construction team member i after standardization.

[0121] The linear regression model of timeliness fitness is fitted using the least squares method and historical data, and the target vector is the historical completion time of each type of task completed by the construction team members.

[0122] The timeliness score of a task reflects its urgency and time constraints. Critical, high-priority tasks typically have higher timeliness scores. For example, a basic construction task like foundation pouring might have a timeliness score of 8 / 10, while some later finishing tasks might have a timeliness score of 5 / 10.

[0123] The Time Adjustment Ability Score calculates each worker's time adjustment ability based on their performance in past projects (e.g., whether they were able to adjust plans on time and complete tasks efficiently). This is measured by factors such as the worker's historical task completion time and the difference from the scheduled time. Standardized Time Adjustment Ability: This standardizes the worker's time adjustment ability and converts it into a standardized score for ease of subsequent calculations.

[0124] When using historical data to fit a linear regression model, the components of the historical data are:

[0125] Member historical completion time: The actual time each construction member took to complete similar tasks. This data reflects the actual time a member spent on a specific task, providing a true time performance indicator for the linear regression model.

[0126] Timeliness score: Each task has been assigned a timeliness score based on factors such as urgency and deadline. The timeliness score helps the model understand the time requirements of the task and is usually between 1 and 10, with higher values indicating more urgent tasks.

[0127] Members' Time Adjustment Ability: This is a member's ability to adjust and manage their time on a task. It is quantified by comprehensively considering a member's experience, past project completion history, and time management skills. Time adjustment ability is one of the key factors in predicting whether a member will complete a task on time.

[0128] This historical data will serve as input to the regression model. By fitting the regression model, the model can learn how to predict a member's timeliness score for a new task based on the task's timeliness score, the member's historical completion time, and their time adjustment ability, thereby helping to better allocate tasks.

[0129] After obtaining the regression coefficient, it is substituted into the linear regression model of timeliness fitness, and the timeliness fitness of construction team members for each type of work task is calculated through the linear regression model of timeliness fitness.

[0130] In one embodiment, using fuzzy logic principles and combining the professional capabilities of construction team members, quantifying the adaptability of construction team members to the authority of each type of work task includes:

[0131] Determine the authoritative score of each category of homework tasks based on the complexity and technical difficulty of each category of homework tasks.

[0132] The academic background and project leadership ability of construction team members are obtained and used as input variables for fuzzy logic reasoning.

[0133] The input variables are converted into fuzzy membership values through membership functions; based on the fuzzy membership values of academic background and project leadership ability, and through all suitable fuzzy inference rules and weighted calculation, the fuzzy reasoning results of construction team members for each type of work task are obtained.

[0134] The defuzzification method is used to convert the fuzzy reasoning results into the authority scores of construction team members for each type of work task.

[0135] The authority score of the construction team members for each type of work task is divided by the authority score of the corresponding type of work task to obtain the adaptability of the construction team members to the authority of each type of work task.

[0136] It should be noted that the authority score of each task is determined based on the technical difficulty and complexity of the task. For example, tasks with higher technical requirements (such as the installation of high-voltage electrical equipment and the erection of complex structural steel frames) will have higher authority scores. In addition to the technical complexity of the task, whether the execution of the task requires a high degree of professional knowledge, the requirements for execution accuracy, and whether it involves the decision-making of the project leader will affect the authority score. Authority score range: defined as a range of 0 to 10, tasks that are highly technical and require high execution accuracy are scored high (such as 9 / 10), and relatively simple tasks are scored low (such as 5 / 10).

[0137] Construction team members' academic background (such as degrees or certifications in relevant fields) is used to assess their adaptability to technical tasks. Generally, workers with higher academic or professional qualifications receive higher academic scores. Project leadership skills are assessed based on project leadership experience, leadership skills, and past project performance.

[0138] Membership functions are used to convert academic background and project leadership skills into fuzzy membership values. Common membership functions include: Academic background membership: For example, member A's academic background (engineering degree) corresponds to a membership value of 0.9, while member B (only a high school diploma) corresponds to a membership value of 0.3. Project leadership membership: For example, member A has extensive experience as a project manager, corresponding to a membership value of 0.8, while member B, who is only an executive, corresponds to a membership value of 0.4.

[0139] According to the rules of fuzzy logic reasoning, the academic background and project leadership ability are combined with the authority scores of certain tasks. For example:

[0140] If a worker has a high academic background and a high project leadership ability, they are more adaptable to high-authority tasks. If a worker has both a low academic background and a low project leadership ability, their adaptability is poor. By weighting the academic background and project leadership ability membership, the fuzzy reasoning results of the member for a certain type of task are calculated. Based on the fuzzy reasoning results, defuzzification methods (such as the center average method and the maximum membership method) are used to convert the fuzzy membership into a specific numerical value. For example, the maximum membership method (i.e., selecting the specific score corresponding to the maximum membership value) is used to determine a member's authority score for a task. For example, if member A's fuzzy reasoning result is 0.85, using the maximum membership method, the member's authority score for the electrical installation task is 8 / 10.

[0141] Member allocation module 3 is used to conduct collaborative analysis on the comprehensive evaluation fitness, criticality fitness, timeliness fitness, authority fitness and social behavior characteristics of construction team members based on graph neural network, and optimize the allocation results of construction team members.

[0142] In one embodiment, based on a graph neural network, a collaborative analysis is performed on comprehensive assessment fitness, criticality fitness, timeliness fitness, authority fitness, and social behavior characteristics of construction team members, and the allocation results of construction team members are optimized, including:

[0143] Obtain the social behavior characteristics of construction team members, including the collaboration efficiency among construction team members.

[0144] Construction team members are regarded as nodes, and the node characteristics are criticality fitness, timeliness fitness, authority fitness and comprehensive evaluation fitness. The edges and edge weights between nodes are determined according to the social behavior characteristics of construction team members.

[0145] A graph neural network structure is constructed, and a message passing mechanism is used to update node features. The node output of the graph neural network is the final feature vector of each construction team member. The final feature vector includes the fitness and social behavior characteristic information of the construction team members for a certain type of construction task after processing by the graph neural network. The fitness includes fitness in terms of criticality, timeliness, authority, and comprehensive evaluation fitness.

[0146] Based on the gap between the fitness of construction team members for a certain type of construction task and the requirements of the corresponding construction task, as well as the collaboration efficiency among construction team members, a loss function is constructed; the allocation of work tasks that can maximize the fitness score and improve the collaboration efficiency of the construction team is taken as the optimization goal, and gradient descent is used to minimize the loss function to obtain a trained graph neural network model.

[0147] Through the trained graph neural network model, the work task requirements are input and the final fitness score of each construction team member is output; the construction team members are allocated according to the final fitness score to optimize the collaboration efficiency among members.

[0148] In one embodiment, the loss function is formulated as:

[0149]

[0150] Where L represents the total value of the loss function; represents the fitness score of construction team member i for task j at the kth iteration, Ta(j) represents the target requirements of task j (such as criticality, timeliness, authority, etc.); wij represents the weighting coefficient of task j to construction team member i, indicating the importance of the task. The determination of the weighting coefficient requires comprehensive consideration of multiple factors such as the criticality, timeliness, complexity, resource requirements, as well as historical execution experience and the judgment of project management experts. Through the reasonable combination of these factors, an appropriate weighting coefficient is determined for each task; λ represents the weight factor of collaboration efficiency; C(S it ) represents the collaboration efficiency between construction team member i and construction team member t; n represents the number of construction team members, and m represents the number of work tasks.

[0151] It should be noted that this invention incorporates multi-dimensional fitness features into the graph neural network, which helps the model more accurately capture the different performance and collaborative efficiency of team members in tasks. This not only enhances the model's ability to model complex tasks and team relationships, but also improves the accuracy of task allocation and team collaboration optimization. By comprehensively considering each member's fitness characteristics in multiple aspects, the network can better allocate tasks, thereby improving the work efficiency of the entire construction team.

[0152] Social behavior traits, such as collaboration efficiency, are indicators of how well team members work together. Collaboration efficiency is assessed based on historical project collaboration performance, such as the quality of completed tasks and the communication and coordination skills of team members. Collaboration efficiency is assigned a value in the range [0, 1], where 0 represents complete non-cooperation and 1 represents complete collaboration.

[0153] In a graph neural network, nodes exchange information with their neighbors through edges. Each node transmits its own features to its neighbors and updates itself based on their features. The updated feature vectors of construction team members contain the criticality, timeliness, authority, and social behavior characteristics processed by the graph neural network.

[0154] The closed-loop management module 4 is used to dynamically adjust the engineering operation mode according to the progress of the engineering operation tasks, combined with the real-time environmental data of the engineering site and the allocation results of the construction team personnel.

[0155] In one embodiment, based on the progress of the engineering task, combined with real-time environmental data at the engineering site and the allocation of construction team personnel, dynamically adjusting the engineering operation mode includes:

[0156] Adjust the allocation of construction team personnel according to the progress of engineering work tasks.

[0157] Adjust the allocation of engineering resources based on real-time environmental data at the engineering site and the progress of engineering tasks.

[0158] Obtain the effect of the adjustment of the engineering operation mode and compare the differences before and after the adjustment until the expected engineering operation progress target is achieved.

[0159] To monitor the progress of engineering tasks, IoT devices or on-site management platforms can be used to track the execution status of each task in real time. This includes task start and completion times, resource consumption, and interruptions. Actual progress can be compared with the planned schedule to calculate the extent of delays or early completion. If the actual progress of a task lags behind schedule, appropriate adjustments should be implemented. IoT devices can also be used to collect real-time on-site environmental data to monitor the impact of weather changes, ambient temperature and humidity, and equipment failures on task execution.

[0160] Dynamically adjust construction team members based on the progress of engineering tasks. If certain tasks are lagging, more members with high timeliness and adaptability can be deployed to accelerate progress. If tasks are ahead of schedule, some members can be transferred to other tasks that require support. Dynamically adjust resource allocation based on real-time environmental data. For example, in extreme weather conditions, construction plans need to be adjusted or protective measures added; in the event of equipment failure, backup equipment needs to be quickly mobilized or repairs arranged. Evaluate the effectiveness of the adjusted operating methods, focusing on changes in task progress, construction quality, worker safety, cost control, and other aspects.

[0161] like Figure 2 According to another embodiment of the present invention, a closed-loop operation management method for sub-projects based on the Internet of Things is provided, comprising:

[0162] S1. Based on the clustering algorithm and the task data collected by the Internet of Things, the task of each part of the project is classified.

[0163] S2. Calculate the construction team members' adaptability to the criticality, timeliness, and authority of each type of task, and integrate the adaptability to obtain a comprehensive assessment of their adaptability.

[0164] S3. Based on graph neural networks, collaborative analysis is conducted on the comprehensive evaluation fitness, criticality fitness, timeliness fitness, authority fitness, and social behavior characteristics of construction team members, and the allocation results of construction team members are optimized.

[0165] S4. Dynamically adjust the engineering operation mode according to the progress of the engineering operation tasks, combined with the real-time environmental data of the engineering site and the allocation results of the construction team personnel.

[0166] In order to facilitate understanding of the above technical solutions of the present invention, the working principle of the present invention in actual process is described in detail below.

[0167] The present invention is applied to large-scale facility construction projects, including but not limited to complex engineering projects such as bridge construction and high-rise building construction.

[0168] For example, in a large-scale bridge construction project, different types of construction tasks need to be categorized and managed. The project is divided into multiple phases, such as foundation construction, steel structure installation, and concrete pouring.

[0169] 1. Task Classification

[0170] (1) Equipment deployment and data collection

[0171] Fifty temperature and humidity sensors were installed at various locations across the construction site, collecting data hourly. Twenty RFID tags were deployed to monitor material consumption in storage areas. Ten video surveillance cameras were installed in key areas to monitor construction progress and safety in real time.

[0172] (2) Data processing and cluster analysis

[0173] The collected data includes a temperature range of 15°C to 30°C, a humidity range of 40% to 70%, and an air quality index fluctuating between 50 and 100. The K-means algorithm was used to classify the work tasks, and the average silhouette coefficient was evaluated by the silhouette coefficient, which was 0.65, indicating that the clustering effect was good. Specific example: The concrete pouring task (environmental requirements: temperature 20°C ± 5°C, humidity 60% ± 10%) was distinguished from other tasks to ensure the quality of task execution under specific environmental conditions. The steel structure installation task (environmental requirements: no strict environmental requirements) was assigned to a separate group.

[0174] 2. Fitness Evaluation Module

[0175] Assess construction team members' suitability for different types of tasks to optimize task allocation.

[0176] For example, for a foundation construction task, a criticality score of 8 / 10 is assigned because it directly impacts the overall project schedule. Team member A has extensive foundation construction experience (professional competence score of 8 / 10) and has a high level of matching experience in similar projects (experience matching score of 9 / 10). Using a weighted scoring method, the criticality fitness is calculated to be 0.9, or (8 + 9) / 10.

[0177] For example, the timeliness score for infrastructure construction tasks is 8 / 10, and member A's historical completion time is 95% of the estimated time (time adjustment ability score is 0.95). Using a linear regression model to calculate timeliness fitness, the formula is: fitness = 0.5 + 0.3 × 8 + 0.2 × 0.95, resulting in a timeliness fitness of 0.85 for member A.

[0178] For example, the authority rating for infrastructure construction tasks is 9 / 10. Team member A holds a master's degree in a related field (academic background membership 0.9) and extensive experience as a project manager (project leadership membership 0.8). Applying fuzzy logic principles and combining membership functions to calculate authority fitness yields a final score of 0.88.

[0179] Set the weight coefficient to:

[0180] Criticality fitness: a1 = 0.5. Timeliness fitness: a2 = 0.3. Authority fitness: a3 = 0.2.

[0181] The two adjustment coefficients are 0.1 and 0.05 respectively.

[0182] The infrastructure construction task's timeliness score is 8 / 10, member A's time management ability is 0.95, and his leadership ability score is 0.85. Substituting these into the formula for comprehensive fitness evaluation, the overall fitness evaluation is 0.45 + 0.105 + 0.175 = 0.73.

[0183] 3. Membership Allocation

[0184] Graph neural networks are used to optimize task allocation among construction team members. The efficiency of collaboration between team members is assessed using historical project collaboration data. For example, if the collaboration efficiency score between team members A and B is 0.85, this indicates that they work well together and can complete tasks efficiently.

[0185] Graph Neural Network Construction and Optimization: All construction team members were treated as nodes, and edge weights were determined based on collaborative efficiency. Assuming a total of 30 members, a 30-node graph neural network was constructed. After multiple iterations (e.g., 50), the loss function value decreased from an initial 150 to a final 30, demonstrating significant optimization results. Ultimately, member A was recommended for infrastructure construction and member B for steel structure installation, given their high fitness scores and excellent collaborative efficiency.

[0186] In task allocation, the loss function is used to measure the gap between the task allocation and the goal, which is to maximize task fitness and improve collaboration efficiency. The loss function consists of two parts:

[0187] Task fitness gap: This part calculates the gap between the task requirements and each member’s fitness, with the goal of ensuring that each member has the best fitness in the task assignment.

[0188] Optimization of collaboration efficiency: This part calculates the collaboration efficiency among construction team members and hopes to maximize the overall collaboration efficiency.

[0189] The optimization goal of the loss function is:

[0190] Minimize the gap between member fitness and task requirements;

[0191] Improve the efficiency of collaboration among members, thereby optimizing the execution effect of the overall team.

[0192] 4. Closed-loop management

[0193] Daily environmental data collection from the construction site revealed that the temperature suddenly soared to 35°C on one day, exceeding the optimal temperature range for concrete pouring. The work plan was adjusted, with additional watering measures added to reduce temperatures. More members with high timeliness and adaptability were deployed to expedite the task.

[0194] Comparing the progress of tasks before and after the adjustment, the concrete pouring task, originally planned to be completed in 7 days, was actually completed in just 6 days, one day ahead of schedule. At the same time, timely cooling measures were implemented to avoid quality issues caused by high temperatures, improving construction quality.

[0195] The present invention comprises:

[0196] (1) Through collaborative analysis using a fitness evaluation matrix and graph neural networks, we comprehensively consider each team member’s capabilities in terms of criticality, timeliness, and authority, as well as the efficiency of collaboration among team members, and optimize the task allocation plan. This precise multi-dimensional evaluation can effectively resolve the mismatch between team member skills and task requirements.

[0197] (2) By collecting on-site data in real time through Internet of Things technology and combining it with data analysis and real-time feedback mechanisms in the closed-loop management module, construction operations can be adjusted dynamically to quickly respond to on-site changes.

[0198] (3) The collaborative analysis model of the graph neural network further improves the collaborative efficiency of the construction team by considering the collaborative efficiency between construction team members and optimizing the collaborative pairing of team members during the task allocation process. This optimization can effectively reduce task conflicts and resource waste, and improve the work coordination and overall execution of the construction team.

[0199] (4) By combining the linear regression model with the weighted average method and combining historical data to evaluate the timeliness adaptability of members, we can ensure that members are reasonably scheduled according to their respective time adjustment capabilities and historical performance, thereby optimizing the execution time of construction tasks and ensuring that the project is completed on schedule.

[0200] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A closed-loop operation management system for sub-projects based on the Internet of Things, characterized by: include: The task classification module is used to classify the work tasks of each part of the project based on clustering algorithms and work task data collected by the Internet of Things; The fitness assessment module is used to calculate the fitness of construction team members in terms of the criticality, timeliness, and authority of each type of work task, and integrate the fitness to obtain a comprehensive fitness assessment; The member allocation module is used to collaboratively analyze the comprehensive evaluation fitness, criticality fitness, timeliness fitness, authority fitness, and social behavior characteristics of construction team members based on graph neural networks, and optimize the allocation results of construction team members; The closed-loop management module is used to dynamically adjust the engineering operation mode according to the progress of the engineering operation tasks, combined with the real-time environmental data of the engineering site and the allocation results of the construction team personnel.

2. The closed-loop operation management system for sub-projects based on the Internet of Things according to claim 1 is characterized in that: The classification of the work tasks of the sub-projects based on the clustering algorithm and the work task data collected by the Internet of Things includes: Plan the installation location of IoT monitoring equipment based on the requirements of the sub-projects at the project construction site; Using IoT monitoring equipment to collect task-related data and extract task features from the task-related data, where the task-related data includes basic task information, task progress information, task feature data, and task equipment information; Based on the clustering algorithm and according to the characteristics of the job tasks, the job tasks are grouped, and the silhouette coefficient is used to evaluate the clustering effect.

3. The closed-loop operation management system for sub-projects based on the Internet of Things according to claim 1 is characterized in that: The calculation of the construction team members' adaptability to the criticality, timeliness, and authority of each type of task, and the integration of adaptability to obtain a comprehensive assessment of adaptability include: Based on the weighted average method and combined with the construction team members' task realization capabilities, the construction team members' adaptability to the critical aspects of each type of work task is calculated; Based on the linear regression model and combined with the time adjustment ability of construction team members, the adaptability of construction team members to the timeliness of each type of task is calculated; By using fuzzy logic principles and combining the professional capabilities of construction team members, we can quantify the adaptability of construction team members to the authority of each type of work task. Based on the construction team members' adaptability to the criticality, timeliness, and authority of each type of task, the comprehensive evaluation adaptability is calculated. The formula for the comprehensive evaluation adaptability is: Where G ij represents the comprehensive evaluation fitness of construction team member i on task j; a1, a2, and a3 represent the importance coefficients of criticality, timeliness, and authority fitness in the comprehensive evaluation of fitness, respectively; δ1 and δ2 are both adjustment coefficients; Key ij represents the fitness of construction team member i in terms of the criticality of task j; Time ij represents the timeliness of construction team member i in task j; Au ij represents the adaptability of construction team member i in terms of authority on task j; T j A represents the timeliness score of task j, ij represents the time adjustment ability of construction team member i on task j, L i represents the project leadership ability score of construction team member i.

4. The closed-loop operation management system for sub-projects based on the Internet of Things according to claim 3 is characterized in that: The calculation of the construction team members' adaptability to the critical aspects of each type of task based on the weighted average method and combined with the construction team members' task realization capabilities includes: Determine the criticality score of each type of task based on its impact on the completion of each part of the project; Obtain the professional capabilities of construction team members associated with each type of work task and their experience matching with each type of work task; Calculate the criticality score of each construction team member for each type of task using a weighted scoring method, combined with the professional ability score and experience matching of the construction team members; The construction team members' criticality scores for each type of work task are divided by the criticality scores of the corresponding type of work tasks to obtain the construction team members' adaptability to the criticality of each type of work task.

5. The closed-loop operation management system for sub-projects based on the Internet of Things according to claim 3 is characterized in that: The calculation of the timeliness adaptability of construction team members to each type of task based on the linear regression model and the time adjustment ability of construction team members includes: Determine the timeliness score of each type of task based on the execution time of each type of task; Obtain the historical completion time and time management ability of construction team members for each type of task, and comprehensively obtain the time adjustment ability of construction team members for each type of task; Normalize the timeliness scores of each type of task and standardize the time adjustment ability; Construct a linear regression model of time-effectiveness fitness: Where, represents the adaptability of construction team member i to the timeliness of task category j; β0 represents the intercept term, β1 and β2 represent the regression coefficients of the timeliness score of the task and the time adjustment ability of the construction team members, respectively; T′ j represents the normalized timeliness score of task category j; C′ i represents the time adjustment ability of construction team member i after standardization; The linear regression model of timeliness fitness is fitted using the least squares method and historical data, and the target vector is the historical completion time of each type of task completed by the construction team members; After obtaining the regression coefficient, it is substituted into the linear regression model of timeliness fitness, and the timeliness fitness of construction team members for each type of work task is calculated through the linear regression model of timeliness fitness.

6. The closed-loop operation management system for sub-projects based on the Internet of Things according to claim 3 is characterized in that: The method of using fuzzy logic principles and combining the professional capabilities of construction team members to quantify the adaptability of construction team members to the authority of each type of task includes: Determine the authority score for each type of task based on its complexity and technical difficulty; Obtain the academic background and project leadership ability of construction team members, and use the academic background and project leadership ability as input variables for fuzzy logic reasoning; The input variables are converted into fuzzy membership values through membership functions. Based on the fuzzy membership values of academic background and project leadership ability, and through all appropriate fuzzy inference rules and weighted calculation, the fuzzy inference results of construction team members for each type of work task are obtained. The defuzzification method is used to convert the fuzzy reasoning results into the authority scores of construction team members for each type of work task; The authority score of the construction team members for each type of work task is divided by the authority score of the corresponding type of work task to obtain the adaptability of the construction team members to the authority of each type of work task.

7. The closed-loop operation management system for sub-projects based on the Internet of Things according to claim 1 is characterized in that: The graph neural network-based collaborative analysis of comprehensive assessment fitness, criticality fitness, timeliness fitness, authority fitness, and social behavior characteristics of construction team members, and the optimization of construction team member allocation results include: Obtaining the social behavior characteristics of construction team members, including the collaboration efficiency among construction team members; Construction team members are regarded as nodes, and the node characteristics are fitness in terms of criticality, timeliness, authority, and comprehensive evaluation. The edges and edge weights between nodes are determined according to the social behavior characteristics of construction team members. Construct a graph neural network structure and use a message passing mechanism to update node features. The node output of the graph neural network is the final feature vector of each construction team member. The final feature vector includes the construction team member's fitness for a certain type of construction task and social behavior characteristic information after being processed by the graph neural network. The fitness includes fitness in terms of criticality, timeliness, authority, and comprehensive evaluation. Based on the gap between the fitness of construction team members for a certain type of construction task and the corresponding construction task requirements, as well as the collaboration efficiency among construction team members, a loss function is constructed. The optimization goal is to allocate work tasks in a way that maximizes the fitness score and improves the collaboration efficiency of the construction team. Gradient descent is used to minimize the loss function, resulting in a trained graph neural network model. Through the trained graph neural network model, the job task requirements are input and the final fitness score of each construction team member is output; the construction team members are assigned according to the final fitness score.

8. The closed-loop operation management system for sub-projects based on the Internet of Things according to claim 7 is characterized in that: The formula of the loss function is: Where L represents the total value of the loss function; represents the fitness score of construction team member i for task j at the kth iteration, and Ta(j) represents the target requirement of task j; w ij represents the weighting coefficient of task j to construction team member i; λ represents the weight factor of collaboration efficiency; C(S it ) represents the collaboration efficiency between construction team member i and construction team member t; n represents the number of construction team members, and m represents the number of work tasks.

9. The closed-loop operation management system for sub-projects based on the Internet of Things according to claim 1 is characterized in that: The dynamic adjustment of the engineering operation mode according to the progress of the engineering operation task, combined with the real-time environmental data of the engineering site and the allocation results of the construction team personnel, includes: Adjust the allocation of construction team personnel according to the progress of engineering tasks; Adjust the allocation of engineering resources based on real-time environmental data at the engineering site and the progress of engineering tasks; Obtain the effect of the adjustment of the engineering operation mode and compare the differences before and after the adjustment until the expected engineering operation progress target is achieved.

10. A closed-loop operation management method for sub-projects based on the Internet of Things, applied to a closed-loop operation management system for sub-projects based on the Internet of Things according to any one of claims 1 to 9, characterized in that: include: Classify the work tasks of each sub-project based on clustering algorithms and work task data collected by the Internet of Things; Calculate the construction team members' adaptability to the criticality, timeliness, and authority of each type of task, and integrate the adaptability to obtain a comprehensive assessment of adaptability; Based on graph neural networks, we conduct collaborative analysis of comprehensive assessment fitness, criticality fitness, timeliness fitness, authority fitness, and the social behavior characteristics of construction team members, and optimize the allocation of construction team members. Dynamically adjust the engineering operation mode according to the progress of the engineering operation tasks, combined with the real-time environmental data of the engineering site and the allocation results of the construction team personnel.

Images