A method for predicting construction project cost based on artificial intelligence

Abstract

A method for predicting construction project cost based on artificial intelligence relates to the technical field of data prediction. The method comprises the following steps: obtaining a total set of construction data and a total set of cost data, setting progress nodes to divide the total set into a subset of construction data and a subset of cost data, obtaining progress coefficients of different progress nodes and a difference set of construction data and a difference value of progress coefficients of adjacent progress nodes, constructing a progress evaluation model, obtaining cost coefficients of different progress nodes and a difference value of cost coefficients of adjacent progress nodes, constructing a cost evaluation model, obtaining a subset of real-time construction data and a difference set of real-time construction data, obtaining a predicted difference value of progress coefficients by using a progress evaluation model, obtaining a predicted difference value of cost coefficients by using a cost evaluation model, and obtaining a predicted cost coefficient by combining the real-time cost coefficient. The method is conducive to realizing an effective prediction mechanism for construction project cost.

Description

Technical Field

[0001] The present invention relates to the technical field of data prediction, and in particular to a method for predicting construction project cost based on artificial intelligence. Background Art

[0002] Cost forecasting for construction projects mainly relies on current artificial intelligence technology to process and analyze a large amount of project cost data, thereby improving the accuracy and efficiency of forecasting. This method combines a variety of artificial intelligence technologies such as deep learning, machine learning, and neural networks to analyze and learn project cost data and build corresponding forecasting models, thereby accurately predicting the cost of subsequent projects.

[0003] Most of the cost forecasts in the prior art are aimed at predicting the total cost at the end of the construction project, but fail to predict the cost at different time points during the construction process. In addition, the cost forecasts of construction projects in the prior art are not combined with the progress forecasts of construction projects, resulting in the prediction results not being able to fit the construction progress of the project, and therefore failing to provide more effective reference value. In view of the shortcomings of the prior art, the present invention provides a construction project cost prediction method based on artificial intelligence. Summary of the invention

[0004] The purpose of the present invention is to provide a construction project cost prediction method based on artificial intelligence.

[0005] The purpose of the present invention can be achieved by the following technical solution: A construction project cost prediction method based on artificial intelligence, comprising the following steps:

[0006] Step S1: obtaining a total set of engineering data and a total set of cost data of different buildings, setting a number of progress nodes, and using the progress nodes to divide the total set of engineering data and the total set of cost data into different engineering data subsets and cost data subsets;

[0007] Step S2: obtaining the progress coefficients of different progress nodes under the same building, and obtaining the engineering data difference set and progress coefficient difference value of adjacent progress nodes, and constructing a progress evaluation model according to the engineering data difference set and progress coefficient difference value of adjacent progress nodes under different buildings;

[0008] Step S3: Obtain the cost coefficients of different progress nodes under different buildings and the cost coefficient differences of adjacent progress nodes, and build a cost assessment model by combining the corresponding engineering data difference set and progress coefficient difference;

[0009] Step S4: obtaining a subset of the real-time engineering data of the current building and a difference set of the real-time engineering data thereof, and obtaining a corresponding predicted progress coefficient difference using a progress assessment model;

[0010] Step S5: Obtain the real-time cost coefficient of the current building, use the cost assessment model to obtain the corresponding predicted cost coefficient difference, and obtain the predicted cost coefficient in combination with the real-time cost coefficient.

[0011] Furthermore, the process of obtaining the total set of engineering data and the total set of cost data of different buildings includes:

[0012] The engineering data set refers to all data related to the construction of a single building from the beginning to the end of the construction, including personnel data, equipment data, material data, and point cloud data;

[0013] The total set of construction cost data refers to various costs related to the construction cost of a single building from the beginning to the end of construction, including labor costs, material costs, equipment costs, and management costs.

[0014] Furthermore, a process of setting a plurality of progress nodes and using the progress nodes to divide the total set of engineering data and the total set of cost data into different engineering data subsets and cost data subsets includes:

[0015] In the engineering data set and the cost data set, several progress nodes are set in sequence according to the time sequence and the preset fixed intervals, the set progress nodes are numbered, and the progress nodes belonging to different buildings but with the same number are bound;

[0016] In the engineering data set, all data before each progress node are included in the engineering data subset of the corresponding progress node; in the cost data set, all data before each progress node are included in the cost data subset of the corresponding progress node.

[0017] Furthermore, the process of obtaining the progress coefficients of different progress nodes under the same building and obtaining the engineering data difference set and progress coefficient difference value of adjacent progress nodes includes:

[0018] Perform 3D modeling based on the latest point cloud data in a single engineering data subset, convert the latest point cloud data into a building model of the corresponding progress node using a voxelization algorithm, and obtain building models of different progress nodes for the same building;

[0019] Obtain design drawings of corresponding buildings, construct theoretical building models according to the design drawings, use the volume ratio between the building model of a single progress node and its theoretical building model as the progress coefficient of the building model of the progress node, and obtain the progress coefficients of different progress nodes under the same building;

[0020] The difference of the progress coefficients of adjacent progress nodes under the same building is taken as the corresponding progress coefficient difference, and the difference set of two engineering data subsets of adjacent progress nodes is taken as the corresponding engineering data difference set, and it is bound to the progress coefficient difference of the corresponding adjacent progress nodes;

[0021] Get the engineering data difference set and progress coefficient difference value of each adjacent progress node under the same building, get the engineering data difference set and progress coefficient difference value of each adjacent progress node under different buildings, and bind the engineering data difference set and progress coefficient difference value of adjacent progress nodes belonging to different buildings but with the same number.

[0022] Furthermore, the process of constructing a progress evaluation model based on the engineering data difference set and progress coefficient difference of adjacent progress nodes under different buildings includes:

[0023] Generate a progress evaluation set according to the engineering data difference set and progress coefficient difference value of adjacent progress nodes with the same number under different buildings, and divide the progress evaluation set into a first training set and a first test set;

[0024] Constructing a first convolutional neural network, using different engineering data difference sets and progress coefficient difference values ​​in the first training set as input data and output data of the first convolutional neural network, respectively, and training the first convolutional neural network to obtain a first initial convolutional neural network;

[0025] The first initial convolutional neural network is model verified using the first test set, and the first initial convolutional neural network with a value less than or equal to a preset first test error threshold is output as a progress assessment model of the building.

[0026] Furthermore, the cost coefficients of different progress nodes under different buildings and the cost coefficient differences of adjacent progress nodes are obtained, and the process of building a cost evaluation model by combining the corresponding engineering data difference set and progress coefficient difference includes:

[0027] The sum of the various costs in the cost data subset of a single progress node is used as the cost coefficient of the progress node, and the cost coefficients of different progress nodes under the same building are obtained, and the difference between the cost coefficients of adjacent progress nodes is used as the cost coefficient difference;

[0028] Obtain the cost coefficient difference of each adjacent progress node under different buildings, generate a cost evaluation set according to the engineering data difference set, progress coefficient difference and cost coefficient difference of adjacent progress nodes with the same number under different buildings, and divide the cost evaluation set into a second training set and a second test set;

[0029] Constructing a second convolutional neural network, using different engineering data difference sets, progress coefficient differences and cost coefficient differences in the second training set as input data and output data of the second convolutional neural network, respectively, and training the second convolutional neural network to obtain a second initial convolutional neural network;

[0030] The second initial convolutional neural network is model verified using the second test set, and a second initial convolutional neural network with a value less than or equal to a preset second test error threshold is output as a cost assessment model for the building.

[0031] Furthermore, the process of obtaining a subset of the real-time engineering data of the current building and a difference set of the real-time engineering data thereof and obtaining a corresponding difference value of the predicted progress coefficient by using the progress evaluation model includes:

[0032] Acquire the real-time engineering data of the current building, which refers to various data related to the construction of the current building recorded since the start of construction, including personnel data, equipment data, material data, and point cloud data;

[0033] In the real-time engineering data, a number of progress nodes are sequentially set in chronological order and at preset fixed intervals, and each data before each progress node is included in the real-time engineering data subset of the corresponding progress node, and the difference between two real-time engineering data subsets of adjacent progress nodes is used as the real-time engineering data difference set;

[0034] The latest real-time engineering data difference set of the current building is input into the progress assessment model, and the progress assessment model is used to output the corresponding predicted progress coefficient difference according to the real-time engineering data difference set.

[0035] Furthermore, the real-time cost coefficient of the current building is obtained, and the corresponding predicted cost coefficient difference is obtained by using the cost evaluation model. The process of obtaining the predicted cost coefficient by combining the real-time cost coefficient includes:

[0036] Obtaining real-time cost data of the current building, which refers to various cost-related data recorded since the start of construction of the current building, including labor cost, material cost, equipment cost, and management cost;

[0037] In the real-time cost data, several progress nodes are sequentially set in chronological order and at preset fixed intervals, and various data before each progress node are included in the real-time cost data subset of the corresponding progress node, and the sum of various costs in the real-time cost data subset is used as the real-time cost coefficient of the corresponding progress node;

[0038] Input the latest real-time engineering data difference set and predicted progress coefficient difference of the current building into the cost assessment model, and use the cost assessment model to output the predicted cost coefficient difference based on the real-time engineering data difference set and predicted progress coefficient difference;

[0039] The sum of the difference between the latest real-time cost coefficient and the predicted cost coefficient is used as the predicted cost coefficient of the next progress node, and the predicted cost coefficient is fed back to relevant personnel.

[0040] Compared with the prior art, the present invention has the following beneficial effects:

[0041] The present invention obtains the total set of engineering data and the total set of cost data of the building, and divides them into different engineering data subsets and cost data subsets according to fixed time intervals, so as to obtain the engineering data difference sets of different progress nodes and their corresponding progress coefficient difference values, and then to construct a corresponding progress evaluation model according to their corresponding relationship, which is conducive to obtaining the corresponding predicted progress coefficient difference value according to the real-time engineering data difference set of the current building;

[0042] By obtaining the cost coefficients of different progress nodes and the cost coefficient differences of adjacent progress nodes, and combining the engineering data difference sets and progress coefficient differences of the corresponding adjacent progress nodes, a building cost assessment model can be constructed, which is beneficial for obtaining the corresponding predicted cost coefficient differences according to the construction progress of the project and real-time engineering data. On the basis of the real-time cost coefficient of the current building, the subsequent predicted cost coefficient can be obtained, which is beneficial for realizing an effective prediction mechanism for the construction project cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 It is a flow chart of the present invention. DETAILED DESCRIPTION

[0044] like Figure 1 As shown, a construction project cost prediction method based on artificial intelligence includes the following steps:

[0045] Step S1: obtaining a total set of engineering data and a total set of cost data of different buildings, setting a number of progress nodes, and using the progress nodes to divide the total set of engineering data and the total set of cost data into different engineering data subsets and cost data subsets;

[0046] Step S2: obtaining the progress coefficients of different progress nodes under the same building, and obtaining the engineering data difference set and progress coefficient difference value of adjacent progress nodes, and constructing a progress evaluation model according to the engineering data difference set and progress coefficient difference value of adjacent progress nodes under different buildings;

[0047] Step S3: Obtain the cost coefficients of different progress nodes under different buildings and the cost coefficient differences of adjacent progress nodes, and build a cost assessment model by combining the corresponding engineering data difference set and progress coefficient difference;

[0048] Step S4: obtaining a subset of the real-time engineering data of the current building and a difference set of the real-time engineering data, and obtaining a corresponding predicted progress coefficient difference using a progress evaluation model;

[0049] Step S5: Obtain the real-time cost coefficient of the current building, use the cost assessment model to obtain the corresponding predicted cost coefficient difference, and obtain the predicted cost coefficient in combination with the real-time cost coefficient.

[0050] It should be further explained that, in the specific implementation process, the process of obtaining the total set of engineering data and the total set of cost data of different buildings includes:

[0051] Taking the construction process of any building as an example, the engineering data set refers to various data related to the construction of the building from the beginning to the end of the construction, including personnel data, equipment data, material data, point cloud data, etc.

[0052] The personnel data refers to the personnel involved in the construction at different times during the construction process, the equipment data refers to the equipment involved in the construction at different times during the construction process, the material data refers to the materials used for construction at different times during the construction process, and the point cloud data refers to the point cloud data of the building collected at different times during the construction process;

[0053] The total set of construction cost data refers to various costs related to the construction cost in the whole process from the beginning to the end of the construction of the building, including labor costs, material costs, equipment costs, management costs, etc.;

[0054] The labor cost refers to the wages, benefits and subsidies of the workers and technicians engaged in the construction, and the material cost refers to the cost of the building materials, equipment, spare parts, hardware and electrical appliances required for the construction, as well as the amortization cost of turnover materials;

[0055] The equipment cost refers to the cost of renting or purchasing machinery, equipment, tools, etc. used in the construction process; the management cost refers to the costs of management staff salaries, travel expenses, office expenses, employee training expenses, etc. incurred in organizing and managing construction.

[0056] It should be further explained that, in the specific implementation process, a number of progress nodes are set, and the process of using the progress nodes to divide the total engineering data set and the total cost data set into different engineering data subsets and cost data subsets includes:

[0057] Each data in the engineering data set and the cost data set has its corresponding recording time. In the engineering data set and the cost data set, a number of progress nodes are set in chronological order, each progress node corresponds to a time node, and the time interval between adjacent progress nodes is a preset fixed interval;

[0058] The set progress nodes are numbered as i, where i=1, 2, ..., n, and n is the number of progress nodes. The number of progress nodes in the engineering data set and the cost data set of different buildings is different. The progress nodes belonging to different buildings but with the same number are bound;

[0059] In the engineering data collection, the data before each progress node are included in the engineering data subset of the corresponding progress node;

[0060] The larger the number of the progress node, the more data its engineering data subset contains, and the engineering data subset of the progress node with a larger number contains all the data in the engineering data subset of the progress node with a smaller number;

[0061] In the total set of cost data, the data before each progress node are included in the cost data subset of the corresponding progress node;

[0062] Among them, the larger the number of the progress node, the more data its cost data subset contains. The cost data subset of the progress node with a larger number contains all the data in the cost data subset of the progress node with a smaller number, and the engineering data subset and cost data subset of the same progress node are bound.

[0063] It should be further explained that, in the specific implementation process, the process of obtaining the progress coefficients of different progress nodes under the same building and obtaining the engineering data difference set and progress coefficient difference of adjacent progress nodes includes:

[0064] Taking the engineering data subset of any progress node as an example, the latest point cloud data in the engineering data subset is divided into different categories, including ground points, building points, road points, etc., and various feature information in the engineering data subset is extracted, including surface normals, edges, three-dimensional structures, etc.;

[0065] Three-dimensional modeling is performed based on the engineering data subset and the extracted feature information. The latest point cloud data is converted into a corresponding three-dimensional model using a voxelization algorithm and marked as the building model of the progress node. The same method is used to obtain the building models of different progress nodes of the same building.

[0066] Obtain design drawings of the corresponding building, and construct a corresponding theoretical building model according to the obtained design drawings, wherein the theoretical building model refers to a theoretical state of the building when the building is completed, and the volume ratio between the building model of a single progress node and its corresponding theoretical building model is used as the progress coefficient of the building model of the progress node;

[0067] The same method is adopted to obtain the progress coefficients of different progress nodes under the same building respectively, and the difference between the progress coefficients of adjacent progress nodes is obtained, which is recorded as the progress coefficient difference. The progress coefficient difference is used to reflect the construction progress change between the building models of adjacent progress nodes;

[0068] Acquire engineering data difference sets of adjacent progress nodes, where the engineering data difference sets refer to the difference sets between two engineering data subsets of adjacent progress nodes, and bind the difference sets to progress coefficient differences of the corresponding adjacent progress nodes;

[0069] The same method is adopted to obtain the engineering data difference set and progress coefficient difference value of each adjacent progress node under the same building, and to obtain the engineering data difference set and progress coefficient difference value of each adjacent progress node under different buildings, and the engineering data difference set and progress coefficient difference value of adjacent progress nodes belonging to different buildings but with the same number are bound.

[0070] It should be further explained that, in the specific implementation process, the process of constructing a progress evaluation model based on the engineering data difference set and progress coefficient difference of adjacent progress nodes under different buildings includes:

[0071] Generate a corresponding progress evaluation set according to the engineering data difference set and progress coefficient difference of adjacent progress nodes with the same number under different buildings, and divide the obtained progress evaluation set into a first training set and a first test set;

[0072] Constructing a first convolutional neural network, using different engineering data difference sets and progress coefficient difference values ​​in the first training set as input data and output data of the first convolutional neural network, respectively, and training the first convolutional neural network to obtain a first initial convolutional neural network;

[0073] The first initial convolutional neural network is model verified using the first test set, and the first initial convolutional neural network with a value less than or equal to a preset first test error threshold is output as a progress assessment model of the building.

[0074] It should be further explained that, in the specific implementation process, the cost coefficients of different progress nodes under different buildings and the cost coefficient differences of adjacent progress nodes are obtained, and the process of building a cost evaluation model by combining the corresponding engineering data difference set and progress coefficient difference includes:

[0075] Taking the cost data subset of any progress node as an example, obtaining the cost coefficient of the progress node, wherein the cost coefficient refers to the sum of various costs in the cost data subset of the progress node;

[0076] Obtain the cost coefficients of different progress nodes under the same building, record the difference between the cost coefficients of adjacent progress nodes as the cost coefficient difference, adopt the same method to obtain the cost coefficient difference of each adjacent progress node under different buildings, and bind the cost coefficient difference of adjacent progress nodes belonging to different buildings but with the same number;

[0077] Generate a corresponding cost evaluation set according to the engineering data difference set, progress coefficient difference value and cost coefficient difference value of adjacent progress nodes with the same number under different buildings, and divide the obtained cost evaluation set into a second training set and a second test set;

[0078] Constructing a second convolutional neural network, using different engineering data difference sets, progress coefficient differences and cost coefficient differences in the second training set as input data and output data of the second convolutional neural network, respectively, and training the second convolutional neural network to obtain a second initial convolutional neural network;

[0079] The second initial convolutional neural network is model verified using the second test set, and a second initial convolutional neural network with a value less than or equal to a preset second test error threshold is output as a cost assessment model for the building.

[0080] It should be further explained that, in the specific implementation process, the process of obtaining the real-time engineering data subset of the current building and its real-time engineering data difference set, and obtaining the corresponding predicted progress coefficient difference using the progress evaluation model includes:

[0081] Acquire the real-time engineering data of the current building, which refers to various data related to the construction of the current building recorded since the start of construction, including personnel data, equipment data, material data, point cloud data, etc.;

[0082] In the real-time engineering data, a number of progress nodes are sequentially set in chronological order and at preset fixed intervals, and each data before each progress node is included in the real-time engineering data subset of the corresponding progress node, and the difference between two real-time engineering data subsets of adjacent progress nodes is used as the real-time engineering data difference set;

[0083] The latest real-time engineering data difference set of the current building is input into the progress assessment model, and the progress assessment model is used to output the corresponding predicted progress coefficient difference according to the real-time engineering data difference set.

[0084] It should be further explained that, in the specific implementation process, the real-time cost coefficient of the current building is obtained, the corresponding predicted cost coefficient difference is obtained by using the cost evaluation model, and the process of obtaining the predicted cost coefficient by combining the real-time cost coefficient includes:

[0085] Obtaining real-time cost data of the current building, which refers to various cost-related data recorded since the start of construction of the current building, including labor cost, material cost, equipment cost, management cost, etc.;

[0086] In the real-time cost data, several progress nodes are sequentially set in chronological order and at preset fixed intervals, and various data before each progress node are included in the real-time cost data subset of the corresponding progress node, and the sum of various costs in the real-time cost data subset is used as the real-time cost coefficient of the corresponding progress node;

[0087] Input the latest real-time engineering data difference set and predicted progress coefficient difference of the current building into the cost assessment model, and use the cost assessment model to output the corresponding predicted cost coefficient difference according to the real-time engineering data difference set and predicted progress coefficient difference;

[0088] The sum of the difference between the latest real-time cost coefficient and the predicted cost coefficient is used as the predicted cost coefficient of the next progress node, and the obtained predicted cost coefficient is fed back to relevant personnel.

[0089] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A construction project cost prediction method based on artificial intelligence, characterized in that: The following steps are involved: Step S1: obtaining a total set of engineering data and a total set of cost data of different buildings, setting a number of progress nodes, and using the progress nodes to divide the total set of engineering data and the total set of cost data into different engineering data subsets and cost data subsets; Step S2: obtaining the progress coefficients of different progress nodes under the same building, and obtaining the engineering data difference set and progress coefficient difference value of adjacent progress nodes, and constructing a progress evaluation model according to the engineering data difference set and progress coefficient difference value of adjacent progress nodes under different buildings; Step S3: Obtain the cost coefficients of different progress nodes under different buildings and the cost coefficient differences of adjacent progress nodes, and build a cost assessment model by combining the corresponding engineering data difference set and progress coefficient difference; Step S4: obtaining a subset of the real-time engineering data of the current building and a difference set of the real-time engineering data, and obtaining a corresponding predicted progress coefficient difference using a progress evaluation model; Step S5: obtaining the real-time cost coefficient of the current building, using the cost assessment model to obtain the corresponding predicted cost coefficient difference, and combining the real-time cost coefficient to obtain the predicted cost coefficient; The process of using the cost assessment model to obtain the difference of the predicted cost coefficient and combining it with the real-time cost coefficient to obtain the predicted cost coefficient includes: Obtaining real-time cost data of the current building, which refers to various cost-related data recorded since the start of construction of the current building, including labor cost, material cost, equipment cost, and management cost; In the real-time cost data, several progress nodes are sequentially set in chronological order and at preset fixed intervals, and various data before each progress node are included in the real-time cost data subset of the corresponding progress node, and the sum of various costs in the real-time cost data subset is used as the real-time cost coefficient of the corresponding progress node; Input the latest real-time engineering data difference set and predicted progress coefficient difference of the current building into the cost assessment model, and use the cost assessment model to output the predicted cost coefficient difference based on the real-time engineering data difference set and predicted progress coefficient difference; The sum of the difference between the latest real-time cost coefficient and the predicted cost coefficient is used as the predicted cost coefficient of the next progress node, and the predicted cost coefficient is fed back to relevant personnel.

2. The method for predicting construction project cost based on artificial intelligence according to claim 1, characterized in that: The process of obtaining the engineering data set and cost data set of different buildings includes: The engineering data set refers to all data related to the construction of a single building from the beginning to the end of the construction, including personnel data, equipment data, material data, and point cloud data; The total set of construction cost data refers to various costs related to the construction cost of a single building from the beginning to the end of construction, including labor costs, material costs, equipment costs, and management costs.

3. The method for predicting construction project cost based on artificial intelligence according to claim 2 is characterized in that: The process of dividing the total engineering data set and the total cost data set into different engineering data subsets and cost data subsets includes: In the engineering data set and the cost data set, several progress nodes are set in sequence according to the time sequence and the preset fixed intervals, the set progress nodes are numbered, and the progress nodes belonging to different buildings but with the same number are bound; In the engineering data set, all data before each progress node are included in the engineering data subset of the corresponding progress node; in the cost data set, all data before each progress node are included in the cost data subset of the corresponding progress node.

4. The method for predicting construction project cost based on artificial intelligence according to claim 3 is characterized in that: The process of obtaining the progress coefficients of different progress nodes and obtaining the engineering data difference set and progress coefficient difference of adjacent progress nodes includes: 3D modeling is performed based on the latest point cloud data in a single engineering data subset, and the latest point cloud data is converted into a building model of the corresponding progress node using a voxelization algorithm to obtain building models of different progress nodes under the same building; Obtain design drawings of corresponding buildings, construct theoretical building models according to the design drawings, use the volume ratio between the building model of a single progress node and its theoretical building model as the progress coefficient of the building model of the progress node, and obtain the progress coefficients of different progress nodes under the same building; The difference of the progress coefficients of adjacent progress nodes under the same building is taken as the corresponding progress coefficient difference, and the difference set of two engineering data subsets of adjacent progress nodes is taken as the corresponding engineering data difference set, and it is bound to the progress coefficient difference of the corresponding adjacent progress nodes; Get the engineering data difference set and progress coefficient difference value of each adjacent progress node under the same building, get the engineering data difference set and progress coefficient difference value of each adjacent progress node under different buildings, and bind the engineering data difference set and progress coefficient difference value of adjacent progress nodes belonging to different buildings but with the same number.

5. The method for predicting construction project cost based on artificial intelligence according to claim 4 is characterized in that: The process of building a progress assessment model includes: Generate a progress evaluation set according to the engineering data difference set and progress coefficient difference value of adjacent progress nodes with the same number under different buildings, and divide the progress evaluation set into a first training set and a first test set; Constructing a first convolutional neural network, using different engineering data difference sets and progress coefficient difference values ​​in the first training set as input data and output data of the first convolutional neural network, respectively, and training the first convolutional neural network to obtain a first initial convolutional neural network; The first initial convolutional neural network is model verified using the first test set, and the first initial convolutional neural network with a value less than or equal to a preset first test error threshold is output as a progress assessment model of the building.

6. The method for predicting construction project cost based on artificial intelligence according to claim 5 is characterized in that: The process of obtaining the cost coefficients of different progress nodes and the cost coefficient differences of adjacent progress nodes and constructing a cost assessment model includes: The sum of the various costs in the cost data subset of a single progress node is used as the cost coefficient of the progress node, and the cost coefficients of different progress nodes under the same building are obtained, and the difference between the cost coefficients of adjacent progress nodes is used as the cost coefficient difference; Obtain the cost coefficient difference of each adjacent progress node under different buildings, generate a cost evaluation set according to the engineering data difference set, progress coefficient difference and cost coefficient difference of adjacent progress nodes with the same number under different buildings, and divide the cost evaluation set into a second training set and a second test set; Constructing a second convolutional neural network, using different engineering data difference sets, progress coefficient differences and cost coefficient differences in the second training set as input data and output data of the second convolutional neural network, respectively, and training the second convolutional neural network to obtain a second initial convolutional neural network; The second initial convolutional neural network is model verified using the second test set, and a second initial convolutional neural network with a value less than or equal to a preset second test error threshold is output as a cost assessment model for the building.

7. The method for predicting construction project cost based on artificial intelligence according to claim 6 is characterized in that: The process of obtaining a subset of real-time engineering data and a difference set of real-time engineering data and obtaining a difference value of a predicted progress coefficient by using a progress evaluation model includes: Acquire the real-time engineering data of the current building, which refers to various data related to the construction of the current building recorded since the start of construction, including personnel data, equipment data, material data, and point cloud data; In the real-time engineering data, a number of progress nodes are sequentially set in chronological order and at preset fixed intervals, and each data before each progress node is included in the real-time engineering data subset of the corresponding progress node, and the difference between two real-time engineering data subsets of adjacent progress nodes is used as the real-time engineering data difference set; The latest real-time engineering data difference set of the current building is input into the progress assessment model, and the progress assessment model is used to output the corresponding predicted progress coefficient difference according to the real-time engineering data difference set.

Images