Big data driven project cost collaborative management system based on BIM (Building Information Modeling)

Abstract

The invention discloses a BIM (Building Information Modeling)-based big data-driven project cost collaborative management system in the technical field of project management, and solves the problem of influence of environmental factors on cost prediction. An environmental influence coefficient A is obtained through calculation, the environmental influence coefficient A and building information model data are unified into an input structure, and the intelligent parameter analysis module realizes significance identification of environmental factors influencing the cost by using deep clustering and linear regression; the system is combined with a dynamic display unit, reflects a cost deviation value delta C in real time and supports multi-path decision analysis, so that a project team can quickly optimize a scheme. Through model optimization training, the self-adaptive engineering cost prediction model improves the estimation accuracy under the assistance of historical data and current characteristics, the budget correction frequency and the financial management complexity are reduced, and the accuracy and effectiveness of engineering cost management are systematically enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of engineering management, and in particular to a big data driven engineering cost collaborative management system based on BIM. Background Art

[0002] The collaborative management system for project cost originated from the construction industry's demand for cost control. With the expansion of project scale and the increase in management complexity, the rapid development of information technology has promoted the digitalization of project cost management and formed a system with collaborative management as its core. Through integrated data sharing, real-time monitoring and multi-party collaboration, the accuracy and transparency of project budgets have been improved, information exchange and efficient cooperation among all project participants have been achieved, and the modernization of project management has been promoted.

[0003] In the existing technology, the publication number is CN119721577A, and its name is a construction cost management system based on big data. This invention involves the field of construction cost technology in the construction industry. The core lies in integrating big data technology to manage construction costs in all aspects. Data management covers collection, storage, analysis and maintenance, ensuring data quality and security, and laying a solid foundation for cost management; accurate cost analysis uses a variety of algorithm models to deeply explore data correlations, achieve accurate predictions and analysis of fluctuation causes, and assist in cost optimization; the efficient business logic layer and user interaction layer operate in coordination, with complete business functions and smooth processes, a friendly interface, strong visualization, and feedback channels to promote optimization; data collection is extensive and storage is efficient, far exceeding traditional solutions; analysis depth and visualization effects are outstanding, providing more valuable decision-making basis; business functions are complete and management is scientific, improving the company's cost control and decision-making capabilities, and user experience is good and security is high.

[0004] However, when the above technical solutions are actually applied to project bidding, environmental factors such as climate change and geological conditions often affect project cost forecasts and actual costs. These factors are usually not fully reflected by conventional BIM data.

[0005] This directly leads to inaccuracies in the cost estimation process. Due to the lack of dynamic tracking of these environmental factors, there may be significant deviations between the initial cost estimate and the actual construction cost.

[0006] Such deviations require multiple budget revisions during project implementation, which increases the complexity of financial management.

[0007] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the Invention

[0008] The purpose of the present invention is to provide a BIM-based big data-driven engineering cost collaborative management system to solve the problems raised in the above background technology.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a BIM-based big data-driven construction cost collaborative management system, comprising:

[0010] A data acquisition and fusion module is used to obtain environmental data and building information model data related to the current project; and intelligently fuse the environmental data and building information model data to generate a first data set;

[0011] An intelligent parameter analysis module is used to perform in-depth cluster analysis on the first dataset and calculate the environmental impact coefficient A to screen out environmental parameters that have a potential impact on project cost changes. This module then combines the building information model data to construct a second dataset, which serves as a common parameter reference for collaborative analysis.

[0012] A historical data comparison module is used to extract historical cost-related data from the second data set and collect current cost-related data for comparison; wherein, a cost deviation value ΔC is calculated using a data analysis algorithm and a deviation summary table is generated;

[0013] The model optimization training module is used to build an adaptive engineering cost prediction model, extract the deviation summary table and common parameters as input, and adjust the weight parameters of the adaptive engineering cost prediction model to obtain the optimized engineering cost prediction results;

[0014] The intelligent decision support module includes a collaborative interactive interface and a multi-party conference interface for integrating the environmental impact coefficient A and project cost forecast results. It also uses data visualization tools to display the impact of cost deviation values on different decision paths. Based on the impact results, it provides online analysis and optimization decisions for the current project construction plan, resource allocation and bidding strategy.

[0015] Furthermore, the environmental data acquisition module specifically includes:

[0016] Environmental sensors deployed at construction sites use IoT technology to automatically collect meteorological data. The sensors read real-time data at predetermined intervals and record it in a data log system.

[0017] By equipping with geological radar and surface deformation monitors, we can regularly remotely extract information on the overall geological changes of the current project; and obtain building information model data including cost-related data, architectural design blueprints, material specifications, and construction plans based on the design platform;

[0018] A machine learning algorithm combining cluster analysis and regression modeling is used for data fusion, and quality inspection is performed during the data fusion process.

[0019] Furthermore, the intelligent parameter analysis module includes a deep cluster analysis unit;

[0020] The deep cluster analysis unit is used to perform unsupervised classification on the standardized data in the first data set, including applying a z-score normalization method to the environmental data in the first data set so that the mean of the original data is zero and the standard deviation is one;

[0021] The K-means clustering algorithm is applied to the standardized data for classification, and the optimal cluster center is found through iteration to finally form the classification results of the environmental data.

[0022] Furthermore, the intelligent parameter analysis module also includes an environmental impact coefficient calculation unit;

[0023] The environmental impact coefficient calculation unit is used to calculate the environmental impact coefficient A for the data category after deep cluster analysis; and to apply the linear regression model to each classified data to establish an impact factor formula. The linear regression formula for obtaining the environmental impact coefficient A is:

[0024]

[0025] Where i represents the index of the environmental impact factor in the classification data sample, and n is the number of all environmental impact factors in the classification data sample;

[0026] A i is the environmental impact coefficient of the i-th environmental impact factor in the classification data sample, X ni represents the value of the nth environmental impact factor in the i-th classification data sample, represents the constant term in the regression model, that is, the baseline deviation of linear regression, b n Represents the regression coefficient corresponding to the nth environmental impact factor in the classification data sample;

[0027] Then, we screen out the environmental parameters that have a significant impact on the change in project cost and quantify them in combination with the environmental impact coefficient, including:

[0028] First, the data categories generated by the deep cluster analysis are screened, the test value t of the linear regression is obtained by calculation, and the statistical significance of each parameter is evaluated.

[0029] By presetting the significance level R and comparing it with the test value t, only the parameters with t greater than R are retained to form a significant parameter set;

[0030] For the significant parameter set, the regression coefficient b is combined with the environmental impact coefficient A to calculate the comprehensive impact weight W ni ;

[0031] Then according to the comprehensive impact weight W ni The specific values of are sorted from high to low, and a weight threshold W is set. Parameters below the weight threshold W are eliminated, and the parameters that are not eliminated are used as the parameter list.

[0032] Furthermore, the historical data comparison module includes a deviation calculation unit;

[0033] The deviation calculation unit extracts historical cost-related data that matches the current project type and time interval from the second data set based on timestamp and project type sorting;

[0034] The mean square error method is used to calculate the difference between the current project cost and the historical project cost data of each group to obtain the cost deviation value ΔC. The calculation formula of the cost deviation value ΔC is:

[0035]

[0036] Among them, h represents the index of the current single item cost data and the historical single item cost data, that is, the hth data point, which is used to match the two sets of data one by one. h Indicates the unit cost of the current data, H h represents the single item cost of historical data, and f represents the total number of two sets of data, that is, the number of data points;

[0037] All calculated cost deviation values ΔC constitute a deviation value set.

[0038] Furthermore, the historical data comparison module also includes a deviation reference generation unit;

[0039] The deviation reference generation unit performs statistical and distribution analysis based on all cost deviation values ΔC in the deviation value set; then determines the overall deviation tendency of the cost data by generating a deviation summary table;

[0040] The deviation summary table is used as a reference for dynamic adjustment of the weights of the adaptive engineering cost prediction model.

[0041] Furthermore, the model optimization training module includes a model construction unit;

[0042] The model building unit is used to build an adaptive engineering cost prediction model with a multi-layer perceptron architecture and process input data;

[0043] The adaptive engineering cost prediction model consists of an input layer, two hidden layers, and an output layer. The input layer receives the historical data feature values and the current project feature values.

[0044] The model weight parameters are initialized using uniform distribution in the range of [-0.1, 0.1];

[0045] The deviation summary table and common parameters are input into the model respectively. The deviation summary table includes the deviation mean, deviation standard deviation and deviation tendency analysis results; while the common parameters include project type, time period and geographical location.

[0046] Furthermore, the model optimization training module includes a weight adjustment unit;

[0047] The weight adjustment unit is used to dynamically adjust the model weight parameters through an optimization algorithm and complete the result integration, wherein the dynamic weight parameter adjustment process is implemented based on the Adam optimization algorithm, and the partial derivatives of the loss function are dynamically updated through the gradient descent method, and the loss function adopts the mean square error;

[0048] The optimization results are verified through test data, and the adjusted weight parameters are applied to the model hidden layer and output layer to form a closed-loop dynamic training framework.

[0049] Furthermore, the intelligent decision support module includes a data integration unit;

[0050] The data integration unit defines the matrix positions of the environmental impact coefficient A and the project cost forecast results, and uses linear algebra tools to merge the matrices; the environmental impact coefficient A and the project cost forecast results are merged through the matrices to form a unified input data structure;

[0051] The environmental impact factor A includes the climatic conditions impact factor, the geological characteristics impact factor, and the construction area restriction factor, and the value range of each influencing factor is quantified as [0,1].

[0052] Furthermore, the intelligent decision support module further includes a dynamic display unit;

[0053] The dynamic display unit is used to display the real-time changes of the cost deviation value ΔC by setting visualization tools such as bar charts and line charts, supporting intuitive analysis of the impact of cost deviation on different decision paths;

[0054] Secondly, by analyzing the impact of cost deviations on construction plans, resource allocation, and bidding strategies, a dynamic decision-making optimization process is carried out, including path analysis, dynamic feedback generation, and decision optimization deployment.

[0055] The specific contents of the dynamic decision optimization process are as follows:

[0056] Path analysis settings:

[0057] The construction plan optimization path, which analyzes the impact of different construction plans on cost deviations; the resource allocation path, which analyzes the impact of dynamic resource allocation, including labor and equipment, on cost deviations; and the bidding path, which analyzes the impact of bidding strategies on cost deviations.

[0058] Dynamic feedback generation, using the impact indicators of deviation values to generate dynamic feedback data:

[0059] Construction plan optimization suggestions: When the cost deviation exceeds a threshold, adjustment models for different construction plans are automatically generated. Resource allocation optimization suggestions: Based on the cost deviation ΔC, resource allocation priorities are generated, suggesting the allocation of resources that have the greatest impact on costs. Bidding strategy suggestions: This includes providing bidding decision-making suggestions, including adjusting bid amounts based on cost deviations.

[0060] Optimize decision-making deployment:

[0061] Online deployment of decision optimization results to the project management system, covering construction plan optimization models, resource allocation plans, and bidding strategy adjustments; after decision optimization deployment, real-time monitoring of bidding results and resource usage is carried out to generate a feedback loop;

[0062] At the same time, the update rate q is preset, and the data visualization tool uses the update rate q to perform real-time updates of the data synchronization mechanism.

[0063] Compared with the prior art, the present invention has the following beneficial effects:

[0064] The present invention systematically solves the problem of potential impact of environmental factors such as climate conditions, geological characteristics and restricted construction areas on project cost prediction and actual costs in engineering projects through the effective combination of data integration units and intelligent parameter analysis modules. The system makes full use of the quantitative calculation of the environmental impact coefficient A, quantifies and matrixes the data of the climate condition influence coefficient, geological characteristic influence coefficient and construction area restriction coefficient, so that environmental factors can be dynamically mapped to the project cost prediction model to form a unified input data structure; by combining deep clustering analysis and linear regression models, the significant environmental parameters that affect the changes in project costs are clarified and standardized, so that the integration of environmental characteristics and building information model data maintains consistency and structural integrity in the data input stage, effectively improving the system's sensitivity and pertinence to potential external environmental impacts.

[0065] The present invention also combines the dynamic display unit of the data visualization tool with the intelligent decision support module to solve the deviation between the initial cost estimate and the actual construction cost caused by the lack of dynamic tracking of environmental factors. Further, through the real-time reflection of the cost deviation value ΔC and the dynamic visualization function for different decision paths, namely construction plan optimization, resource allocation analysis and bidding strategy optimization, the engineering project team can adjust the plan more quickly and accurately, effectively reducing the frequency of multiple budget revisions during project execution; through the preset update rate q to ensure real-time synchronization of data, combined with the adaptive engineering cost prediction model of the model optimization training module, through the multi-layer perceptron architecture including historical data features and current project features, based on the dynamically adjusted weight parameters, the accuracy of the current project cost prediction is improved, thereby reducing the complexity of financial management and enhancing budget controllability, ensuring the excellent performance of the system under complex changes in environmental factors. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] 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 or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0067] Figure 1 It is a schematic diagram of the overall system framework structure of the present invention. DETAILED DESCRIPTION

[0068] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only 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 making creative efforts are within the scope of protection of the present invention.

[0069] Example 1

[0070] See also Figure 1 The present invention provides a technical solution: a BIM-based big data-driven engineering cost collaborative management system, including:

[0071] A data acquisition and fusion module is used to obtain environmental data and building information model data related to the current project; and intelligently fuse the environmental data and building information model data to generate a first data set;

[0072] An intelligent parameter analysis module is used to perform in-depth cluster analysis on the first dataset and calculate the environmental impact coefficient A to screen out environmental parameters that have a potential impact on project cost changes. This module then combines the building information model data to construct a second dataset, which serves as a common parameter reference for collaborative analysis.

[0073] A historical data comparison module is used to extract historical cost-related data from the second data set and collect current cost-related data for comparison; wherein, a cost deviation value ΔC is calculated using a data analysis algorithm and a deviation summary table is generated;

[0074] The model optimization training module is used to build an adaptive engineering cost prediction model, extract the deviation summary table and common parameters as input, and adjust the weight parameters of the adaptive engineering cost prediction model to obtain the optimized engineering cost prediction results;

[0075] The intelligent decision support module includes a collaborative interactive interface and a multi-party conference interface for integrating the environmental impact coefficient A and project cost forecast results. It also uses data visualization tools to display the impact of cost deviation values on different decision paths. Based on the impact results, it provides online analysis and optimization decisions for the current project construction plan, resource allocation and bidding strategy.

[0076] In this embodiment, the data acquisition and fusion module generates a first data set by intelligently fusing environmental data and building information model data, thereby ensuring the consistency and integrity of data input and improving sensitivity to potential external influencing factors;

[0077] The intelligent parameter analysis module uses deep cluster analysis and calculation of the environmental impact coefficient (AA) to screen out environmental parameters with potential impact on project cost changes. Combined with the building information model data, it reconstructs a second data set, providing an effective reference for collaborative analysis and enhancing the system's adaptability to environmental changes.

[0078] The historical data comparison module extracts and compares historical cost-related data with current cost data, uses data analysis algorithms to accurately calculate the cost deviation value ΔC, and forms a deviation summary table, thereby improving the understanding of cost change trends and supporting accurate decision-making;

[0079] The model optimization training module builds an adaptive engineering cost prediction model, uses the deviation summary table and common parameters as input, adjusts the model weight parameters, and effectively improves the prediction accuracy and model adaptability;

[0080] The intelligent decision support module integrates the environmental impact coefficient A and the optimized project cost forecast results, and uses data visualization tools to demonstrate the impact of cost deviation values on different decision paths. It supports the project team's online analysis and optimization decisions on construction plans, resource allocation and bidding strategies, thereby improving decision-making efficiency and response speed.

[0081] Example 2

[0082] The environmental data acquisition module specifically includes:

[0083] Environmental sensors deployed at construction sites use IoT technology to automatically collect meteorological data. The sensors read real-time data at predetermined intervals and record it in a data log system.

[0084] By equipping with geological radar and surface deformation monitors, we can regularly remotely extract information on the overall geological changes of the current project; and obtain building information model data including cost-related data, architectural design blueprints, material specifications, and construction plans based on the design platform;

[0085] A machine learning algorithm combining cluster analysis and regression modeling is used for data fusion, and quality inspection is performed during the data fusion process.

[0086] Furthermore, IoT sensors enable automated data acquisition, ensuring timeliness and accuracy, reducing manual intervention, and improving data collection efficiency. Geological monitoring equipment can provide high-precision data, safeguarding the geological safety of engineering projects and closely related to construction planning. Automated updates to BIM model data help improve data consistency and accuracy, making it easier for construction and management teams to keep abreast of project progress. Machine learning technology provides effective multi-data source integration capabilities, making data easier to analyze and providing data support for subsequent engineering cost management modules.

[0087] Example 3

[0088] The intelligent parameter analysis module includes a deep cluster analysis unit;

[0089] The deep cluster analysis unit is used to perform unsupervised classification on the standardized data in the first data set, including applying a z-score normalization method to the environmental data in the first data set so that the mean of the original data is zero and the standard deviation is one;

[0090] Furthermore, the standardization formula for environmental data is:

[0091]

[0092] Where Z represents the environmental data, X is the original data value, μ is the mean of the data set, and σ is the standard deviation of the data set;

[0093] The K-means clustering algorithm is applied to the standardized data for classification, and the optimal cluster center is found through iteration to finally form the classification results of the environmental data.

[0094] The intelligent parameter analysis module also includes an environmental impact coefficient calculation unit;

[0095] The environmental impact coefficient calculation unit is used to calculate the environmental impact coefficient A for the data category after deep cluster analysis; and to apply the linear regression model to each classified data to establish an impact factor formula. The linear regression formula for obtaining the environmental impact coefficient A is:

[0096]

[0097] Where i represents the index of the environmental impact factor in the classification data sample, and n is the number of all environmental impact factors in the classification data sample;

[0098] A i is the environmental impact coefficient of the i-th environmental impact factor in the classification data sample, X ni represents the value of the nth environmental impact factor in the i-th classification data sample, represents the constant term in the regression model, that is, the baseline deviation of linear regression, b n Represents the regression coefficient corresponding to the nth environmental impact factor in the classification data sample;

[0099] Then, we screen out the environmental parameters that have a significant impact on the change in project cost and quantify them in combination with the environmental impact coefficient, including:

[0100] First, the data categories generated by the deep cluster analysis are screened, the test value t of the linear regression is obtained by calculation, and the statistical significance of each parameter is evaluated.

[0101] Furthermore, the calculation formula of the test value t is: ,in is the regression coefficient Standard error of

[0102] By presetting the significance level R and comparing it with the test value t, only the parameters with t greater than R are retained to form a significant parameter set;

[0103] For the significant parameter set, the regression coefficient b is combined with the environmental impact coefficient A to calculate the comprehensive impact weight W ni ;

[0104] Furthermore, the comprehensive impact weight W ni The calculation formula is: ;

[0105] Then according to the comprehensive impact weight W ni The specific values of are sorted from high to low, and a weight threshold W is set. Parameters below the weight threshold W are eliminated, and the parameters that are not eliminated are used as the parameter list.

[0106] The historical data comparison module includes a deviation calculation unit;

[0107] The deviation calculation unit extracts historical cost-related data that matches the current project type and time interval from the second data set based on timestamp and project type sorting;

[0108] The mean square error method is used to calculate the difference between the current project cost and the historical project cost data of each group to obtain the cost deviation value ΔC. The calculation formula of the cost deviation value ΔC is:

[0109]

[0110] Among them, h represents the index of the current single item cost data and the historical single item cost data, that is, the hth data point, which is used to match the two sets of data one by one. h Indicates the unit cost of the current data, H h represents the single item cost of historical data, and f represents the total number of two sets of data, that is, the number of data points;

[0111] All calculated cost deviation values ΔC constitute a deviation value set.

[0112] The historical data comparison module also includes a deviation reference generation unit;

[0113] The deviation reference generation unit performs statistical and distribution analysis based on all cost deviation values ΔC in the deviation value set; then determines the overall deviation tendency of the cost data by generating a deviation summary table;

[0114] The deviation summary table is used as a reference for dynamic adjustment of the weights of the adaptive engineering cost prediction model.

[0115] In this embodiment, the intelligent parameter analysis module introduces a deep clustering analysis unit and an environmental impact coefficient calculation unit. Different from the existing technology, it processes environmental data using unsupervised classification combined with the z-score normalization method to achieve data standardization and find the optimal cluster center through K-means clustering to form the environmental data classification results, ensuring the accuracy and applicability of the environmental data.

[0116] The environmental impact coefficient calculation unit further performs linear regression analysis on these classification results, calculates the environmental impact coefficient A, and screens out significant impact parameters through the test value t to form a significant parameter set, combined with the comprehensive impact weight W niSorting and screening are performed, which is significantly different from the shortcomings of existing technologies that cannot dynamically quantify environmental factors;

[0117] The deviation calculation unit in the historical data comparison module extracts matching historical data by sorting based on timestamps and project types, calculates the cost deviation value ΔC, and forms a deviation value set to ensure that cost assessment is based on continuous analysis of actual data changes;

[0118] The deviation reference generation unit uses the deviation value set to generate a deviation summary table to guide the model weight adjustment and avoid the misleading caused by static analysis. Through the collection and calculation and evaluation of these parameters, the module realizes the closed-loop functions of data accuracy, dynamic quantification and decision support, effectively improving the dynamic response capability and decision-making accuracy of engineering cost management.

[0119] Example 4

[0120] The model optimization training module includes a model construction unit;

[0121] The model building unit is used to build an adaptive engineering cost prediction model with a multi-layer perceptron architecture and process input data;

[0122] The adaptive engineering cost prediction model consists of an input layer, two hidden layers, and an output layer. The input layer receives the historical data feature values and the current project feature values.

[0123] The model weight parameters are initialized using uniform distribution in the range of [-0.1, 0.1];

[0124] The deviation summary table and common parameters are input into the model respectively. The deviation summary table includes the deviation mean, deviation standard deviation and deviation tendency analysis results; while the common parameters include project type, time period and geographical location.

[0125] Furthermore, the mean deviation is the average of the deviation values, the standard deviation is the degree of dispersion of the deviation values, and the deviation tendency analysis results are used to reflect the trend characteristics of the data, that is, the rate of change of increase or decrease year by year;

[0126] Project type refers to residential and commercial buildings, project time period refers to the duration of the project construction phase, and geographical location refers to the geographical characteristics of the project location;

[0127] Input bias summary table and common parameters to the input layer of the prediction model for training

[0128] The deviation summary table can fully reflect the historical performance of the data, and the public parameters provide auxiliary features for each day, project type, and location to ensure the accuracy of model training;

[0129] The model optimization training module includes a weight adjustment unit;

[0130] The weight adjustment unit is used to dynamically adjust the model weight parameters through an optimization algorithm and complete the result integration, wherein the dynamic weight parameter adjustment process is implemented based on the Adam optimization algorithm, and the partial derivatives of the loss function are dynamically updated through the gradient descent method, and the loss function adopts the mean square error;

[0131] Furthermore, the weight update formula is:

[0132]

[0133] in, represents the current weight, Represents the "weight value in the next iteration" in the dynamic weight update formula, that is, the optimized model parameter weight;

[0134] η is the learning rate, which specifies the step size during gradient descent and controls the amplitude of weight update, and is set to 0.001; L represents the loss function. Represents the partial derivative of the loss function with respect to the current weight parameter.

[0135] Further, The specific role of is to update the weight value during the gradient descent process and calculate the weight of the next model.

[0136] and The specific role of is the starting point of the optimization formula, indicating the weight parameters that have not been updated in the current model;

[0137] It is used to determine the descending direction of the weight parameter on the loss function curve at the current moment, calculate the gradient in this direction through partial derivatives, and guide how the weight should be adjusted to minimize the value of the loss function L. By gradually iteratively updating the weight, the model can eventually converge to the weight parameter with the best prediction accuracy;

[0138] η is used to avoid slow training due to a small step size, or ineffective weight convergence due to a large step size; the loss function L is a function that reflects the error between the predicted value and the true value, and is defined by the mean square error MSE. The specific formula is: Where, is the predicted value, is the true value, v represents the total number of samples in the data set, and e represents the index of the sample data, which is defined as the e-th sample;

[0139] The optimization results are verified through test data, and the adjusted weight parameters are applied to the model hidden layer and output layer to form a closed-loop dynamic training framework.

[0140] The specific process for verifying the optimization results is as follows: First, the dynamically adjusted weight parameters are applied to the hidden and output layers of the adaptive construction cost prediction model. Subsequently, the model is validated using an independent test dataset. This involves using the model to predict the test data and calculating the deviation between the predicted results and the true values. This is typically done through validation metrics such as mean squared error (MSE) to evaluate the model's predictive accuracy and performance. This step verifies whether the optimized model has good generalization capabilities on new data. The model's performance is also recorded and analyzed to improve and ensure a closed-loop model training process, enabling it to adaptively adjust to continuously updated input data and maintain the highest possible prediction accuracy.

[0141] Example 5

[0142] The intelligent decision support module includes a data integration unit;

[0143] The data integration unit defines the matrix positions of the environmental impact coefficient A and the project cost forecast results, and uses linear algebra tools to merge the matrices; the environmental impact coefficient A and the project cost forecast results are merged through the matrices to form a unified input data structure;

[0144] The environmental impact factor A includes the climatic conditions impact factor, the geological characteristics impact factor, and the construction area restriction factor, and the value range of each influencing factor is quantified as [0,1].

[0145] Furthermore, the rainfall frequency Aa and the ratio of construction interruption days Ab are extracted and dimensionlessly processed to obtain the climate condition influence coefficient A1. The specific calculation formula is as follows:

[0146]

[0147] Among them, the construction interruption days ratio Ab represents the proportion of construction interruption days caused by extreme weather (such as typhoons, storms, etc.) to the total construction period;

[0148] Furthermore, the proportion of foundation treatment costs Ac and the unit cost of rock excavation Ad are extracted and dimensionless, and the geological characteristics influence coefficient A2 is calculated. The specific calculation formula is as follows:

[0149]

[0150] Among them, the proportion of foundation treatment costs Ac clarifies the economic impact of foundation treatment on stability; the unit cost of rock excavation Ad is used to quantify the impact of rock excavation difficulty on construction cost.

[0151] Furthermore, the site utilization rate Ae and the proportion of additional equipment costs Af are extracted and dimensionless processed to calculate the construction area restriction coefficient A3. The specific calculation formula is as follows:

[0152]

[0153] Among them, the site utilization rate Ae represents the ratio of the actual available site area to the ideal site area; the equipment entry additional cost ratio Af represents the ratio of the additional cost of large equipment entry due to restricted access to the total construction cost

[0154] Furthermore, the format of the input data structure is:

[0155]

[0156] Among them, A is the environmental impact coefficient vector of 1×u, and P is the cost forecast result vector of 1×y; among them, u represents the number of all environmental factors that have potential impact on the project cost, and the specific number depends on the project requirements; y represents the number of cost forecast results at different construction stages, and the specific number depends on the scale and complexity of the project.

[0157] The intelligent decision support module also includes a dynamic display unit;

[0158] The dynamic display unit is used to display the real-time changes of the cost deviation value ΔC by setting visualization tools such as bar charts and line charts, supporting intuitive analysis of the impact of cost deviation on different decision paths;

[0159] Secondly, by analyzing the impact of cost deviations on construction plans, resource allocation, and bidding strategies, a dynamic decision-making optimization process is carried out, including path analysis, dynamic feedback generation, and decision optimization deployment.

[0160] The specific contents of the dynamic decision optimization process are as follows:

[0161] Path analysis settings:

[0162] The construction plan optimization path, which analyzes the impact of different construction plans on cost deviations; the resource allocation path, which analyzes the impact of dynamic resource allocation, including labor and equipment, on cost deviations; and the bidding path, which analyzes the impact of bidding strategies on cost deviations.

[0163] Dynamic feedback generation, using the impact indicators of deviation values to generate dynamic feedback data:

[0164] Construction plan optimization suggestions: When the cost deviation exceeds a threshold, adjustment models for different construction plans are automatically generated. Resource allocation optimization suggestions: Based on the cost deviation ΔC, resource allocation priorities are generated, suggesting the allocation of resources that have the greatest impact on costs. Bidding strategy suggestions: This includes providing bidding decision-making suggestions, including adjusting bid amounts based on cost deviations.

[0165] Optimize decision-making deployment:

[0166] Online deployment of decision optimization results to the project management system, covering construction plan optimization models, resource allocation plans, and bidding strategy adjustments; after decision optimization deployment, real-time monitoring of bidding results and resource usage is carried out to generate a feedback loop;

[0167] At the same time, the update rate q is preset, and the data visualization tool uses the update rate q to perform real-time updates of the data synchronization mechanism.

[0168] Further, the functional description and technical logic are progressive:

[0169] The data integration unit defines the quantitative calculation of the environmental impact coefficient A and the project cost forecast result P, the integration of input data, and the formation of a unified data structure as the basis for subsequent analysis and presentation. Data input must maintain consistency and structural integrity to ensure that subsequent presentation tools can interpret the data.

[0170] Based on the data integration unit, the dynamic display unit further limits the design of the dynamic visualization display unit's real-time visualization function of cost deviation values and decision path analysis tools; it graphically displays the unified data structure and supports multi-dimensional engineering decision path selection through the real-time analysis module.

[0171] The significant difference between the intelligent decision support module and existing technologies is that it proposes clear quantitative rules and matrix data structures for environmental impact coefficients through the data integration unit, solving the problem of insufficient data input consistency in existing technologies.

[0172] By combining dynamic real-time graphical display with decision-path impact analysis tools, the system can provide real-time feedback on data results. Its real-time decision-making capabilities in engineering projects are significantly superior to existing static analysis solutions.

[0173] The following are specific implementation examples of dynamic decision optimization in construction projects:

[0174] Enter the data structure through the data integration unit:

[0175] Environmental factor A includes the influence factor of climate conditions, the influence factor of geological characteristics and the construction area restriction factor;

[0176] The current estimated total project cost is RMB 1.35 million;

[0177] Table 1 is the input data structure table:

[0178]

[0179] Real-time display through dynamic display units combined with visualization tools:

[0180] Table 2 is a bar chart showing the cost deviation value ΔC:

[0181]

[0182] The dynamic decision optimization process includes:

[0183] 1. Path analysis settings:

[0184] Path analysis for optimizing construction plans: Analyze the impact of new materials and construction schedules, temperature, and the impact of adjustments that slow concrete curing time on cost;

[0185] Path analysis of resource allocation paths: labor and equipment utilization, including optimization of labor allocation during peak hours;

[0186] Path analysis of bidding paths: optimal bidding solution, taking into account the market competitiveness of the target project.

[0187] 2. Dynamic feedback generation:

[0188] Construction plan optimization suggestions: If the cost deviation exceeds 5%, it is recommended to adjust the construction schedule and material selection;

[0189] Table 3 is the conditional structure table of the construction plan optimization suggestion:

[0190]

[0191] Generate resource priority based on cost deviation value ΔC;

[0192] Table 4 is a table of resource allocation optimization suggestions:

[0193]

[0194] Bidding strategy recommendations: generate bidding decision recommendations and adjust bid amounts;

[0195] Table 5 is a table of bidding strategy recommendations:

[0196]

[0197] Optimize decision-making deployment:

[0198] Online deployment: Input optimization decisions into the project management system, covering construction plan optimization models, resource allocation plans, and bidding strategy adjustments.

[0199] Real-time monitoring and feedback loop: The project management system continuously monitors bidding results and resource utilization.

[0200] Effect evaluation:

[0201] Cost control: Achieve cost deviation below target value and keep it within 1.5%.

[0202] Resource optimization: Improve resource utilization by approximately 20% and reduce labor costs by 10%.

[0203] Improved efficiency: The adjusted construction plan increased the project execution rate by approximately 12%.

[0204] It should be noted that: All calculation formulas in this application document use regression analysis including but not limited to machine learning algorithms to deeply analyze the relevant parameters collected and identify their natural trends and relationships. Use professional software, such as Python's Scikit-learn library or R language, to automatically generate mathematical models that match the data. Then, objectively evaluate the performance of the model through methods such as cross-validation, and combine continuous feedback and optimization to ensure that the created formula truly reflects the inherent laws of the data, thereby ensuring its effectiveness and accuracy. In all calculation formulas in this application, the parameters in each formula are dimensionlessly processed in a consistent range to ensure that different physical quantities are compared on the same scale; dimensionless technical means include but are not limited to Min-Max Normalization and Z-Score standardization;

[0205] The technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as a computer floppy disk, read-only memory (ROM), random access memory (RAM), flash memory (FLASH), hard disk or optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute the methods of various embodiments of the present invention.

[0206] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0207] The above embodiments are only intended to help understand the method and core concept of the present invention. It should be noted that, without departing from the principles of the present invention, a number of improvements and modifications may be made to the present invention by those skilled in the art, and such improvements and modifications also fall within the scope of protection of the claims of the present invention.

Claims

1. A BIM-based big data-driven construction cost collaborative management system, characterized by: include: Data acquisition and fusion module, used to obtain environmental data and building information model data related to the current project; and intelligently integrate the environmental data and the building information model data to generate a first data set; An intelligent parameter analysis module is used to perform in-depth cluster analysis on the first dataset and calculate the environmental impact coefficient A to screen out environmental parameters that have a potential impact on project cost changes. This module then combines the building information model data to construct a second dataset, which serves as a common parameter reference for collaborative analysis. A historical data comparison module is used to extract historical cost-related data from the second data set and collect current cost-related data for comparison; wherein, a cost deviation value ΔC is calculated using a data analysis algorithm and a deviation summary table is generated; The model optimization training module is used to build an adaptive engineering cost prediction model, extract the deviation summary table and common parameters as input, and adjust the weight parameters of the adaptive engineering cost prediction model to obtain the optimized engineering cost prediction results; The intelligent decision support module includes a collaborative interactive interface and a multi-party conference interface for integrating the environmental impact coefficient A and project cost forecast results. It also uses data visualization tools to display the impact of cost deviation values on different decision paths. Based on the impact results, it provides online analysis and optimization decisions for the current project construction plan, resource allocation and bidding strategy.

2. The BIM-based big data-driven construction cost collaborative management system according to claim 1 is characterized by: The environmental data acquisition module specifically includes: Environmental sensors deployed at construction sites use IoT technology to automatically collect meteorological data. The sensors read real-time data at predetermined intervals and record it in a data log system. By equipping with geological radar and surface deformation monitors, we can regularly remotely extract information on the overall geological changes of the current project; and obtain building information model data including cost-related data, architectural design blueprints, material specifications, and construction plans based on the design platform; A machine learning algorithm combining cluster analysis and regression modeling is used for data fusion, and quality inspection is performed during the data fusion process.

3. The BIM-based big data-driven construction cost collaborative management system according to claim 1 is characterized by: The intelligent parameter analysis module includes a deep cluster analysis unit; The deep cluster analysis unit is used to perform unsupervised classification on the standardized data in the first data set, including applying a z-score normalization method to the environmental data in the first data set so that the mean of the original data is zero and the standard deviation is one; The K-means clustering algorithm is applied to the standardized data for classification, and the optimal cluster center is found through iteration to finally form the classification results of the environmental data.

4. The BIM-based big data-driven construction cost collaborative management system according to claim 1 is characterized by: The intelligent parameter analysis module also includes an environmental impact coefficient calculation unit; The environmental impact coefficient calculation unit is used to calculate the environmental impact coefficient A for the data category after deep cluster analysis; and to apply the linear regression model to each classified data to establish an impact factor formula. The linear regression formula for obtaining the environmental impact coefficient A is: ； Where i represents the index of the environmental impact factor in the classification data sample, and n is the number of all environmental impact factors in the classification data sample; A i is the environmental impact coefficient of the i-th environmental impact factor in the classification data sample, X ni represents the value of the nth environmental impact factor in the i-th classification data sample, represents the constant term in the regression model, that is, the baseline deviation of linear regression, b n Represents the regression coefficient corresponding to the nth environmental impact factor in the classification data sample; Then, we screen out the environmental parameters that have a significant impact on the change in project cost and quantify them in combination with the environmental impact coefficient, including: First, the data categories generated by the deep cluster analysis are screened, the test value t of the linear regression is obtained by calculation, and the statistical significance of each parameter is evaluated. By presetting the significance level R and comparing it with the test value t, only the parameters with t greater than R are retained to form a significant parameter set; For the significant parameter set, the regression coefficient b is combined with the environmental impact coefficient A to calculate the comprehensive impact weight W ni ; Then according to the comprehensive impact weight W ni The specific values of are sorted from high to low, and a weight threshold W is set. Parameters below the weight threshold W are eliminated, and the parameters that are not eliminated are used as the parameter list.

5. The BIM-based big data-driven construction cost collaborative management system according to claim 4 is characterized by: The historical data comparison module includes a deviation calculation unit; The deviation calculation unit extracts historical cost-related data that matches the current project type and time interval from the second data set based on timestamp and project type sorting; The mean square error method is used to calculate the difference between the current project cost and the historical project cost data of each group to obtain the cost deviation value ΔC. The calculation formula of the cost deviation value ΔC is: ； Among them, h represents the index of the current single item cost data and the historical single item cost data, that is, the hth data point, which is used to match the two sets of data one by one. h Indicates the unit cost of the current data, H h represents the single item cost of historical data, and f represents the total number of two sets of data, that is, the number of data points; All calculated cost deviation values ΔC constitute a deviation value set.

6. The BIM-based big data-driven construction cost collaborative management system according to claim 5 is characterized by: The historical data comparison module also includes a deviation reference generation unit; The deviation reference generation unit performs statistical and distribution analysis based on all cost deviation values ΔC in the deviation value set; then determines the overall deviation tendency of the cost data by generating a deviation summary table; The deviation summary table is used as a reference for dynamic adjustment of the weights of the adaptive engineering cost prediction model.

7. The BIM-based big data-driven construction cost collaborative management system according to claim 6 is characterized by: The model optimization training module includes a model construction unit; The model building unit is used to build an adaptive engineering cost prediction model with a multi-layer perceptron architecture and process input data; The adaptive engineering cost prediction model consists of an input layer, two hidden layers, and an output layer. The input layer receives the historical data feature values and the current project feature values. The model weight parameters are initialized using uniform distribution; The deviation summary table and common parameters are input into the model respectively. The deviation summary table includes the deviation mean, deviation standard deviation and deviation tendency analysis results; while the common parameters include project type, time period and geographical location.

8. The BIM-based big data-driven construction cost collaborative management system according to claim 7 is characterized by: The model optimization training module includes a weight adjustment unit; The weight adjustment unit is used to dynamically adjust the model weight parameters through an optimization algorithm and complete the result integration, wherein the dynamic weight parameter adjustment process is implemented based on the Adam optimization algorithm, and the partial derivatives of the loss function are dynamically updated through the gradient descent method, and the loss function adopts the mean square error; The optimization results are verified through test data, and the adjusted weight parameters are applied to the model hidden layer and output layer to form a closed-loop dynamic training framework.

9. The BIM-based big data-driven construction cost collaborative management system according to claim 8, characterized in that: The intelligent decision support module includes a data integration unit; The data integration unit defines the matrix positions of the environmental impact coefficient A and the project cost forecast results, and uses linear algebra tools to merge the matrices; the environmental impact coefficient A and the project cost forecast results are merged through the matrices to form a unified input data structure; The environmental impact coefficient A includes the climatic conditions influence coefficient, the geological characteristics influence coefficient and the construction area restriction coefficient, and the value range of each influencing factor is quantified as [0,1].

10. The BIM-based big data-driven construction cost collaborative management system according to claim 9 is characterized by: The intelligent decision support module also includes a dynamic display unit; The dynamic display unit is used to display the real-time changes of the cost deviation value ΔC by setting visualization tools such as bar charts and line charts, supporting intuitive analysis of the impact of cost deviation on different decision paths; Secondly, by analyzing the impact of cost deviations on construction plans, resource allocation, and bidding strategies, a dynamic decision-making optimization process is carried out, including path analysis, dynamic feedback generation, and decision optimization deployment. The specific contents of the dynamic decision optimization process are as follows: Path analysis settings: The construction plan optimization path, which analyzes the impact of different construction plans on cost deviations; the resource allocation path, which analyzes the impact of dynamic resource allocation, including labor and equipment, on cost deviations; and the bidding path, which analyzes the impact of bidding strategies on cost deviations. Dynamic feedback generation, using the impact indicators of deviation values to generate dynamic feedback data: Construction plan optimization suggestions: When the cost deviation exceeds a threshold, adjustment models for different construction plans are automatically generated. Resource allocation optimization suggestions: Based on the cost deviation ΔC, resource allocation priorities are generated, suggesting the allocation of resources that have the greatest impact on costs. Bidding strategy suggestions: This includes providing bidding decision-making suggestions, including adjusting bid amounts based on cost deviations. Optimize decision-making deployment: Online deployment of decision optimization results to the project management system, covering construction plan optimization models, resource allocation plans, and bidding strategy adjustments; after decision optimization deployment, real-time monitoring of bidding results and resource usage is carried out to generate a feedback loop; At the same time, the update rate q is preset, and the data visualization tool uses the update rate q to perform real-time updates of the data synchronization mechanism.

Images