Intelligent construction supervision system based on BIM and intelligent algorithm

Abstract

The invention discloses an intelligent construction supervision system based on BIM and an intelligent algorithm, and the system comprises a BIM modeling module which builds a building information model according to project plan data through the BIM technology; the collection module is used for collecting task data corresponding to J tasks of the construction project, and the task data comprises task feature data, construction environment features and worker feature data; the task feature data comprises a task type, task difficulty and a task scale; the task scale is determined according to a building information model; the analysis module is used for inputting the task data corresponding to the J tasks into the task time calculation model and outputting time required by the tasks corresponding to the J tasks; and efficient configuration and reasonable arrangement of tasks are realized, and the construction delay risk and cost are effectively reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of construction supervision, and in particular to an intelligent construction supervision system based on BIM and intelligent algorithms. Background Art

[0002] Currently, although some project management software uses BIM technology for visual management, with the rapid development of the construction industry, the challenges faced by construction project management are becoming increasingly complex, and these management tools still have significant limitations.

[0003] Construction projects usually involve multiple interdependent tasks, and the relationships between these tasks may be extremely complex and difficult to manage effectively through traditional methods. The dependencies between tasks may cause delays in a task, which directly affects the execution of subsequent tasks and thus affects the progress of the entire construction project. However, traditional project management methods often fail to fully consider these complex task relationships, resulting in time management and construction efficiency being restricted.

[0004] In the traditional construction supervision process, project planning often relies on experience and manual calculations, which is not only prone to human errors, but also difficult to respond quickly to complex task relationships and spatial conflicts. For example, when a task is delayed, the supervisor needs to re-evaluate the progress and resource allocation of all related tasks, which is a cumbersome and inefficient process; and the time utilization of workers is often not fully optimized. This leads to construction delays and rising costs.

[0005] Therefore, it is an urgent problem for technical personnel in this field to propose an intelligent construction supervision system based on BIM and intelligent algorithms to solve the difficulties existing in the prior art. Summary of the invention

[0006] The purpose of the present invention is to provide an intelligent construction supervision system based on BIM and intelligent algorithms to solve the shortcomings of the background technology.

[0007] In order to achieve the above-mentioned object, the present invention provides the following technical solutions: an intelligent construction supervision system based on BIM and intelligent algorithms, comprising: a BIM modeling module, which uses BIM technology to establish a building information model according to construction project planning data;

[0008] The collection module is used to collect task data corresponding to J tasks of the construction project, and the task data includes task characteristic data, construction environment characteristic data and worker characteristic data; the task characteristic data includes task type, task difficulty and task scale; the task scale is determined according to the building information model;

[0009] An analysis module, used for inputting task data corresponding to J tasks into a task time calculation model, and outputting task times corresponding to the J tasks;

[0010] The constraint setting module is used to set construction constraints according to the construction project. The construction constraints include worker uniqueness constraints, task worker demand constraints, task duration constraints, task sequence constraints and space conflict constraints.

[0011] The goal setting module is used to set construction goals, including minimizing the completion time of all tasks and minimizing the idle time of workers;

[0012] The construction optimization module is used to select the optimal task execution sequence based on construction constraints and construction goals, relying on the natural optimization heuristic algorithm.

[0013] Furthermore, the worker uniqueness constraint is that each worker can only perform at most one task at any time; the task worker demand constraint is that the number of workers for each task is not less than the number of workers required for the task; the task sequence constraint is that a certain task can only be started after a certain task is completed; the space conflict constraint is that tasks in the same space are carried out simultaneously without conflicting with each other.

[0014] Furthermore, according to the construction constraints and construction goals, the method of selecting the optimal task execution sequence based on the natural optimization heuristic algorithm includes:

[0015] Step 101, set the population size to N, and randomly generate N task sequence sequences;

[0016] Step 102, calculating the fitness value of each task sequence in the population;

[0017] Step 103, sort the fitness values ​​in descending order, select the task sequence corresponding to the first k fitness values, k<N, and mark them as the parent generation task sequence;

[0018] Step 104: randomly select W groups of parent task sequence sequences, W < k, perform crossover operation on two parent task sequence sequences in one group with random exchange intervals to generate child task sequence sequences, and determine whether there are missing or repeated tasks in the child task sequence sequences. If there are, repair them so that each task in the child task sequence sequence appears once;

[0019] Step 105: Perform random position mutation on the tasks in the task sequence sequence of the child generation, generate a new task sequence sequence, and update the population; the random position mutation includes random position swapping, random position merging, or random position splitting;

[0020] Step 106, repeatedly execute steps 102 to 105, and stop iteration when the preset number of iterations is reached, select the task sequence with the highest fitness, and obtain the optimal task execution sequence.

[0021] Furthermore, the method for calculating the fitness value includes:

[0022]

[0023] Where D is the completion time of all tasks in each task sequence; P is the penalty value of each task sequence; H is the sum of the idle time of each worker after completing each task sequence, that is, the sum of the time when each worker is not assigned a task; λ1, λ2, and λ4 are all preset coefficients.

[0024] Furthermore, the penalty value calculation method includes:

[0025] P=λ1P 工人 +λ2P 先后 +λ3P 空间 ;

[0026] P 工人 is the total number of violations of the worker uniqueness constraint; P 先后 is the number of task pairs that violate the task order constraint; P 空间 is the number of task pairs that violate the spatial conflict constraint, and λ1, λ2, and λ3 are all preset coefficients.

[0027] Furthermore, the task types include civil engineering, electrical engineering, water supply and drainage, and decoration; the difficulty of the task is quantified by numerical values; the task scale determines the area and volume of the corresponding task according to the constructed building information model;

[0028] The worker characteristic data include the number of workers corresponding to each task, the unique identifier of each worker and the efficiency evaluation value of each worker; the construction environment characteristics include the weather conditions, working environment and season corresponding to each task; the weather conditions are the number of sunny days, rainy days, snowy days, high temperature days, low temperature days and normal temperature days during the construction phase of the corresponding task, and the weather conditions in the future time period are counted through the weather forecast website; the working environment includes indoors, outdoors or high altitude; the season includes spring, summer, autumn or winter.

[0029] Furthermore, the training method of the task time calculation model includes:

[0030] Collect b groups of task data in advance, set the corresponding task time for each of the b groups of task data, b is an integer greater than 1, and convert the task data and the corresponding task time into a corresponding set of feature vectors;

[0031] Each group of feature vectors is used as the input of a task time calculation model. The task time calculation model uses a group of predicted task times corresponding to each group of task data as output, and uses the actual task time corresponding to each group of task data as a prediction target; minimizing the sum of prediction errors of all task data is used as a training target; the task time calculation model is trained until the sum of prediction errors reaches convergence, and the training is stopped; the task time calculation model is a multilayer perceptron or a long short-term memory network.

[0032] Furthermore, the method for obtaining the worker efficiency evaluation value includes:

[0033] Inputting each worker's historical worker data into a pre-built worker efficiency evaluation model to obtain an efficiency evaluation value corresponding to each worker;

[0034] Worker data include worker characteristics, worker task characteristics, worker construction environment characteristics, and worker health characteristics;

[0035] Worker characteristics include unique identifier, age, gender, length of service and historical performance score. The historical performance score is the worker's absenteeism rate after the completion of the last project. Worker task characteristics include the corresponding task type, task difficulty and task scale at the end of the last task. Worker construction environment characteristics include the corresponding weather conditions, working environment and season at the end of the last task. Worker task characteristics are obtained through statistics at the end of the last task. Worker health characteristics include good, average or poor.

[0036] Furthermore, the training method of the worker efficiency evaluation model includes:

[0037] Collect y groups of historical worker data in advance, set corresponding efficiency evaluation values ​​for the y groups of historical worker data, y is an integer greater than 1, and convert the historical worker data and the corresponding efficiency evaluation values ​​into a corresponding set of feature vectors;

[0038] Each group of feature vectors is used as the input of a worker efficiency evaluation model. The worker efficiency evaluation model uses a group of predicted efficiency evaluation values ​​corresponding to each group of historical worker data as output, and uses the actual efficiency evaluation values ​​corresponding to each group of historical worker data as prediction targets. Minimizing the sum of prediction errors of all historical worker data is used as a training target. The worker efficiency evaluation model is trained until the sum of prediction errors reaches convergence, and the training is stopped. The worker efficiency evaluation model is linear regression, random forest regression, or gradient boosting regression.

[0039] Furthermore, the weather conditions in future time periods announced by real-time weather forecast websites are monitored, and the number of sunny days, rainy days, snowy days, high temperature days, low temperature days and normal temperature days in each task construction phase are counted in real time, and the task time corresponding to J tasks is updated in real time. The optimal task execution sequence is updated in real time based on the natural optimization-inspired algorithm.

[0040] The present invention provides the following technical effects and advantages of an intelligent construction supervision system based on BIM and intelligent algorithms:

[0041] Based on the project planning data, a building information model of the integrated building structure is constructed through the three-dimensional BIM model, and a system of optimal task execution sequence is constructed through intelligent algorithms, providing strong support for construction project supervision. The system can accurately predict the time required for the task by collecting detailed data for each task, and set the constraints and goals of the project on this basis, including multiple constraints such as worker uniqueness, worker demand, task duration, task sequence, and spatial conflicts, to ensure the rationality of the construction plan and reduce human errors.

[0042] The optimization goal of the project is to minimize the total project duration and workers' idle time, so as to effectively shorten the construction period and improve the utilization rate of workers' working time. In terms of optimization algorithms, natural heuristic algorithms are introduced, especially genetic algorithms to select the optimal task execution sequence. In the genetic algorithm, random position mutation operations are added, and further expanded to random position merging and splitting strategies to adapt to the complex needs of construction projects. This improvement enables the system to flexibly handle tasks that can be performed simultaneously in the same space without causing conflicts, thereby achieving maximum compression of the construction period. At the same time, reducing workers' idle time not only improves construction efficiency, but also increases workers' income security.

[0043] Through the intelligent construction supervision system, project managers can achieve efficient configuration and reasonable arrangement of tasks while ensuring project quality, effectively reducing the risk and cost of construction delays, and laying the foundation for intelligent and refined construction management. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0045] Figure 1 It is a schematic diagram of the intelligent construction supervision system based on BIM and intelligent algorithm of the present invention. DETAILED DESCRIPTION

[0046] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0047] like Figure 1 As shown, the intelligent construction supervision system based on BIM and intelligent algorithm in this embodiment includes a BIM modeling module, a collection module, an analysis module, a constraint setting module, a goal setting module and a construction optimization module, which are connected by wire and / or wirelessly.

[0048] BIM modeling module uses BIM technology to establish a building information model based on construction project planning data.

[0049] The collection module is used to collect task data corresponding to J tasks of the construction project, where J is an integer greater than 1. The task data includes task feature data, construction environment features and worker feature data; the task feature data includes task type, task difficulty and task scale; task types include civil engineering, electrical engineering, water supply and drainage, decoration, etc.; the task difficulty is quantified with a numerical value, which is determined by the project leader based on the specific project. The greater the task difficulty, the larger the numerical value; the task scale determines the area and volume of the corresponding task based on the pre-built BIM (Building Information Model), and is expressed numerically.

[0050] Worker characteristic data includes the number of workers corresponding to each task (preset value), each worker's unique identifier and each worker's efficiency evaluation value; construction environment characteristics include weather conditions, working environment and seasons corresponding to each task. Weather conditions refer to the number of sunny days, rainy days, snowy days, high temperature days, low temperature days and normal temperature days during the construction phase of the corresponding task, which are specifically counted through the weather conditions in the future time period announced by the weather forecast website. Working environments include indoors, outdoors, high altitudes, etc.; seasons include spring, summer, autumn and winter.

[0051] The analysis module is used to input the task data corresponding to the J tasks into the task time calculation model, and output the task time corresponding to the J tasks.

[0052] The training method of the task time calculation model includes:

[0053] Collect b groups of task data in advance, set the corresponding task time for each of the b groups of task data, b is an integer greater than 1, and convert the task data and the corresponding task time into a corresponding set of feature vectors.

[0054] Each group of feature vectors is used as the input of a task time calculation model. The task time calculation model takes a group of predicted task time corresponding to each group of task data as output, and takes the actual task time corresponding to each group of task data as the prediction target; minimizing the sum of prediction errors of all task data is used as the training target; the task time calculation model is trained until the sum of prediction errors reaches convergence, and the training is stopped; the task time calculation model is a multi-layer perceptron (MLP) and a long short-term memory network (LSTM).

[0055] Methods for obtaining worker efficiency evaluation values ​​include:

[0056] The historical worker data of each worker is input into the pre-built worker efficiency evaluation model to obtain the efficiency evaluation value corresponding to each worker.

[0057] Worker data include worker characteristics, worker task characteristics, worker construction environment characteristics and worker health characteristics.

[0058] Worker characteristics include unique identifier, age, gender, length of service and historical performance score. The historical performance score is the percentage of absences after the worker's last project was completed, which is used to reflect the worker's work enthusiasm. Worker task characteristics include the corresponding task type, task difficulty and task scale at the end of the last task. Worker construction environment characteristics include the corresponding weather conditions, working environment and season at the end of the last task. Worker task characteristics are obtained by statistics at the end of the last task. Worker health characteristics include good, average or poor, which are obtained through the physical examination report of each worker.

[0059] The training method of the worker efficiency evaluation model includes:

[0060] Collect y groups of historical worker data in advance, set corresponding efficiency evaluation values ​​for each of the y groups of historical worker data, where y is an integer greater than 1, and convert the historical worker data and the corresponding efficiency evaluation values ​​into a corresponding set of feature vectors.

[0061] Each group of feature vectors is used as the input of a worker efficiency evaluation model. The worker efficiency evaluation model takes a group of predicted efficiency evaluation values ​​corresponding to each group of historical worker data as output, and takes the actual efficiency evaluation value corresponding to each group of historical worker data as a prediction target; minimizing the sum of prediction errors of all historical worker data is used as a training target; the worker efficiency evaluation model is trained until the sum of prediction errors reaches convergence, and the training is stopped; the worker efficiency evaluation model is linear regression, random forest regression or gradient boosting regression (GBR).

[0062] The constraint setting module is used to set construction constraints according to the construction project. The construction constraints include worker uniqueness constraints, task worker demand constraints, task duration constraints, task sequence constraints, and space conflict constraints.

[0063] The worker uniqueness constraint is that at any time, each worker can only perform one task at most; the task worker demand constraint is that the number of workers for each task is not less than the number of workers required for the task; the task precedence constraint is that a task can only start after a certain task is completed, as shown in Table 1. The spatial conflict constraint is that tasks in the same space can be carried out simultaneously without conflicting with each other, for example, the pouring and masonry construction of beams, columns, and slabs can be carried out simultaneously with the pre-buried pipelines and cable laying, and there will be no conflict.

[0064]

[0065] Table 1

[0066] The goal setting module is used to set construction goals. The construction goals include minimizing the completion time of all tasks (i.e. minimizing the sum of the time required for all tasks) and minimizing the idle time of workers (i.e. minimizing the time when workers are not assigned to work).

[0067] The construction optimization module is used to select the optimal task execution sequence based on construction constraints and construction goals, relying on the natural optimization heuristic algorithm.

[0068] According to the construction constraints and construction goals, the methods of selecting the optimal task execution sequence based on the natural optimization heuristic algorithm include:

[0069] Step 101, set the population size to N, randomly generate N task order sequences, and N is 50-100.

[0070] Step 102: Calculate the fitness value of each task sequence in the population.

[0071]

[0072] Where D is the completion time of all tasks in the corresponding task sequence; P is the penalty value of the corresponding task sequence, which indicates the degree of violation of the construction constraint. The larger the value, the more severe the penalty; H is the sum of the idle time of each worker after the completion of the corresponding task sequence, that is, the sum of the time when each worker is not assigned a task.

[0073] P=λ1P 工人 +λ2P 先后 +λ3P 空间 ;

[0074] P 工人 is the total number of violations of the worker uniqueness constraint; P 先后 is the number of task pairs that violate the task order constraint;; P 空间is the number of task pairs that violate the spatial conflict constraint; λ1, λ2, and λ3 are preset coefficients used to adjust the weights of various penalties.

[0075] Step 103, sort the fitness values ​​in descending order, select the task sequence corresponding to the first k fitness values, k<N, and mark them as the parent task sequence. The task sequence can also be selected in a "roulette" manner.

[0076] Step 104, randomly select W groups of parent task sequence sequences, W<k, perform crossover operation on two parent task sequence sequences in one group with random exchange intervals, generate child task sequence sequences, and determine whether there are missing or repeated tasks in the child task sequence sequences. If there are, repair them so that each task in the child task sequence sequence appears once; examples of two parent task sequence sequences are as follows:

[0077] {F1, F2, F3, F4, F5, F6F7};

[0078] {F5, F4, F3, F2, F1, F6, F7};

[0079] For example, if the exchange interval is 2-4, then the 2nd to 4th positions in the first parent task sequence: F2, F3, F4 are cross-operated with the 2nd to 4th positions in the first parent task sequence: F4, F3, F2. The generated child task sequence is as follows:

[0080] {F1, F4, F3, F2, F5, F6F7};

[0081] {F5, F2, F3, F4, F1, F6, F7}.

[0082] Step 105: Perform random position mutation on the tasks in the descendant task sequence to generate a new task sequence and update the population; the random position mutation includes random position swapping, random position merging or random position splitting.

[0083] An example of random position swapping is as follows:

[0084] Atomic generation task order sequence {F1, F2, F3, F4, F5, F6F7}; random position: the second and fifth positions; the new task order sequence after position mutation: {F1, F5, F3, F4, F2, F6F7}.

[0085] Random position merging, for example, merging F5 and F4 into F5F4. Random position splitting, for example, splitting F6F7 into F6 and F7. In the genetic algorithm, random position mutation adds the operation of random position merging or random position splitting. This mutation strategy can help cope with the complex needs of construction projects, especially for handling some tasks that can be performed simultaneously in the same space without conflict. This operation can shorten the total project duration to the greatest extent by arranging tasks reasonably, and reduce the idle time of workers, thereby improving the income security of workers.

[0086] It should be noted that random position merging: if some tasks can be performed simultaneously in the same space without interfering with each other, these tasks are merged and scheduled to be performed in the same time period to optimize space utilization. This operation is particularly suitable for task combinations that do not interfere with each other.

[0087] Random position splitting: For some tasks that have potential conflicts but are not obvious, split the tasks into multiple subtasks and spread them over different time periods to avoid mutual interference, thus ensuring the smooth and safe progress of the project.

[0088] Step 106, repeatedly execute steps 102 to 105, and stop iteration when the preset number of iterations is reached, select the task sequence with the highest fitness, and obtain the optimal task execution sequence.

[0089] This example constructs a three-dimensional BIM model that includes building structure, task sequence and resource requirements based on the construction project plan data as the basis of the entire system. It collects the task data corresponding to each task and predicts the time required for each task. It sets construction constraints and construction goals, such as worker uniqueness constraints, task worker demand constraints, task duration constraints, task sequence constraints, and spatial conflict constraints. Construction goals include minimizing the completion time of all tasks and minimizing the idle time of workers. According to the construction constraints and construction goals, the optimal task execution sequence is selected based on the natural optimization heuristic algorithm.

[0090] In the genetic algorithm, the random position mutation is added with the operation of random position merging or random position splitting. This mutation strategy can help cope with the complex demands in construction projects, especially dealing with some tasks that can be performed simultaneously in the same space without conflict. This operation can shorten the total construction period of the project to the greatest extent and reduce the workers' idle time by reasonably arranging tasks, thereby improving the workers' income security.

[0091] Furthermore, the weather conditions in future time periods announced by real-time weather forecast websites are monitored, and the number of sunny days, rainy days, snowy days, high temperature days, low temperature days and normal temperature days in each task construction phase are counted, and the task time corresponding to J tasks is updated in real time. The optimal task execution sequence is updated in real time based on the natural optimization-inspired algorithm; ensuring that the optimal task execution sequence is matched in real time according to the changing data of each task during the entire construction phase.

[0092] By real-time monitoring of future weather conditions provided by weather forecast websites, we systematically count the number of sunny days, rainy days, snowy days, high temperature days, low temperature days, and normal temperature days in each construction phase. Based on these data, we update the construction time required for each task in real time, and dynamically adjust the optimal task execution sequence based on the natural optimization heuristic algorithm.

[0093] It can not only optimize the construction progress, but also effectively reduce the risk of construction delays caused by weather changes. By matching the change data of each task in real time, the project ensures the stable execution of the entire construction phase, can quickly respond to sudden weather changes, and improve the safety and efficiency of construction.

[0094] Intelligent construction supervision has established a full-process management system for the construction process through data-driven methods, providing stable guarantees for the smooth progress of construction projects. It is not only applicable to current projects, but also sets a reference standard for intelligent supervision and optimized management of similar projects in the future, achieving more scientific decision-making and more efficient resource allocation.

[0095] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. Intelligent construction supervision system based on BIM and intelligent algorithm, characterized by: include: BIM modeling module, which uses BIM technology to build a building information model based on the construction project plan data; The collection module is used to collect task data corresponding to J tasks of the construction project, and the task data includes task characteristic data, construction environment characteristic data and worker characteristic data; the task characteristic data includes task type, task difficulty and task scale; the task scale is determined according to the building information model; An analysis module, used for inputting task data corresponding to J tasks into a task time calculation model, and outputting the task time corresponding to the J tasks; The constraint setting module is used to set construction constraints according to the construction project. The construction constraints include worker uniqueness constraints, task worker demand constraints, task duration constraints, task sequence constraints and space conflict constraints. The goal setting module is used to set construction goals, including minimizing the completion time of all tasks and minimizing the idle time of workers; The construction optimization module is used to select the optimal task execution sequence based on construction constraints and construction goals, relying on the natural optimization heuristic algorithm.

2. The intelligent construction supervision system based on BIM and intelligent algorithm according to claim 1 is characterized in that: The worker uniqueness constraint is that each worker can only perform at most one task at any time; the task worker demand constraint is that the number of workers for each task is not less than the number of workers required for the task; the task sequence constraint is that a certain task can only be started after a certain task is completed; the space conflict constraint is that tasks in the same space are carried out simultaneously without conflicting with each other.

3. The intelligent construction supervision system based on BIM and intelligent algorithm according to claim 1 is characterized in that: According to the construction constraints and construction goals, the methods of selecting the optimal task execution sequence based on the natural optimization heuristic algorithm include: Step 101, set the population size to N, and randomly generate N task sequence sequences; Step 102, calculating the fitness value of each task sequence in the population; Step 103, sort the fitness values ​​in descending order, select the task sequence corresponding to the first k fitness values, k<N, and mark them as the parent generation task sequence; Step 104: randomly select W groups of parent task sequence sequences, W < k, perform crossover operation on two parent task sequence sequences in one group with random exchange intervals to generate child task sequence sequences, and determine whether there are missing or repeated tasks in the child task sequence sequences. If there are, repair them so that each task in the child task sequence sequence appears once; Step 105: Perform random position mutation on the tasks in the task sequence sequence of the child generation, generate a new task sequence sequence, and update the population; the random position mutation includes random position swapping, random position merging, or random position splitting; Step 106, repeatedly execute steps 102 to 105, and stop iteration when the preset number of iterations is reached, select the task sequence with the highest fitness, and obtain the optimal task execution sequence.

4. The intelligent construction supervision system based on BIM and intelligent algorithm according to claim 3 is characterized in that: The method for calculating the fitness value includes: Where D is the completion time of all tasks in each task sequence; P is the penalty value of each task sequence; H is the sum of the idle time of each worker after completing each task sequence, that is, the sum of the time when each worker is not assigned a task; λ1, λ2, and λ4 are all preset coefficients.

5. The intelligent construction supervision system based on BIM and intelligent algorithm according to claim 4 is characterized in that: The penalty value calculation method includes: P=λ1P 工人 +λ2P 先后 +λ3P 空间 ; P 工人 is the total number of violations of the worker uniqueness constraint; P 先后 is the number of task pairs that violate the task order constraint; P 空间 is the number of task pairs that violate the spatial conflict constraint, and λ1, λ2, and λ3 are all preset coefficients.

6. The intelligent construction supervision system based on BIM and intelligent algorithm according to claim 1 is characterized in that: The task types include civil engineering, electrical engineering, water supply and drainage, and decoration; the difficulty of the task is quantified by numerical values; the task scale is determined by the area and volume of the corresponding task according to the constructed building information model; The worker characteristic data include the number of workers corresponding to each task, the unique identifier of each worker and the efficiency evaluation value of each worker; the construction environment characteristics include the weather conditions, working environment and season corresponding to each task; the weather conditions are the number of sunny days, rainy days, snowy days, high temperature days, low temperature days and normal temperature days during the construction phase of the corresponding task, and the weather conditions in the future time period are counted through the weather forecast website; the working environment includes indoors, outdoors or high altitude; the season includes spring, summer, autumn or winter.

7. The intelligent construction supervision system based on BIM and intelligent algorithm according to claim 6 is characterized in that: The training method of the task time calculation model includes: Collect b groups of task data in advance, set the corresponding task time for each of the b groups of task data, b is an integer greater than 1, and convert the task data and the corresponding task time into a corresponding set of feature vectors; Each group of feature vectors is used as the input of a task time calculation model. The task time calculation model uses a group of predicted task times corresponding to each group of task data as output, and uses the actual task time corresponding to each group of task data as a prediction target; minimizing the sum of prediction errors of all task data is used as a training target; the task time calculation model is trained until the sum of prediction errors reaches convergence, and the training is stopped; the task time calculation model is a multilayer perceptron or a long short-term memory network.

8. The intelligent construction supervision system based on BIM and intelligent algorithm according to claim 6 is characterized in that: The method for obtaining the worker efficiency evaluation value includes: Inputting each worker's historical worker data into a pre-built worker efficiency evaluation model to obtain an efficiency evaluation value corresponding to each worker; Worker data include worker characteristics, worker task characteristics, worker construction environment characteristics, and worker health characteristics; Worker characteristics include unique identifier, age, gender, length of service and historical performance score. The historical performance score is the worker's absenteeism rate after the completion of the last project. Worker task characteristics include the corresponding task type, task difficulty and task scale at the end of the last task. Worker construction environment characteristics include the corresponding weather conditions, working environment and season at the end of the last task. Worker task characteristics are obtained through statistics at the end of the last task. Worker health characteristics include good, average or poor.

9. The intelligent construction supervision system based on BIM and intelligent algorithm according to claim 8 is characterized in that: The training method of the worker efficiency evaluation model includes: Collect y groups of historical worker data in advance, set corresponding efficiency evaluation values ​​for the y groups of historical worker data, y is an integer greater than 1, and convert the historical worker data and the corresponding efficiency evaluation values ​​into a corresponding set of feature vectors; Each group of feature vectors is used as the input of a worker efficiency evaluation model. The worker efficiency evaluation model uses a group of predicted efficiency evaluation values ​​corresponding to each group of historical worker data as output, and uses the actual efficiency evaluation values ​​corresponding to each group of historical worker data as prediction targets. Minimizing the sum of prediction errors of all historical worker data is used as a training target. The worker efficiency evaluation model is trained until the sum of prediction errors reaches convergence, and the training is stopped. The worker efficiency evaluation model is linear regression, random forest regression, or gradient boosting regression.

10. The intelligent construction supervision system based on BIM and intelligent algorithm according to claim 6 is characterized in that: Monitor the weather conditions in the future time period announced by the real-time weather forecast website, and make real-time statistics on the number of sunny days, rainy days, snowy days, high temperature days, low temperature days and normal temperature days in each task construction phase, update the task time corresponding to J tasks in real time, and update the optimal task execution sequence in real time based on the natural optimization inspired algorithm.

Images