Building engineering cost analysis method and system based on big data analysis
By establishing a construction project cost analysis model through big data analysis and transfer learning, the problem of low efficiency in existing technologies has been solved, and efficient and accurate cost analysis has been achieved.
Patent Information
- Application Number
- CN202511027941.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-24
- Publication Date
- 2025-11-21
AI Technical Summary
Current technologies for analyzing construction project costs are inefficient, requiring highly experienced personnel for accurate analysis, and are time-consuming and inefficient.
By employing a big data analytics approach, an initial construction project cost analysis model is obtained. Then, using transfer learning to train a small number of construction project cost analysis training samples, a target construction project cost analysis model is established to conduct efficient and accurate cost analysis.
It enables efficient and accurate acquisition of construction project cost analysis results, reduces training time and resource consumption, and improves analysis efficiency.
Smart Images
Figure CN120996881A_ABST
Abstract
Description
Technical Field
[0001] This application relates to the technical field of cost analysis in construction engineering, and in particular to a method and system for cost analysis of construction engineering based on big data analysis. Background Technology
[0002] Construction cost management plays a crucial role in engineering construction. It permeates the entire construction process, from the decision-making, design, bidding, and construction stages to the final settlement stage, all of which require precise control of construction costs. In the project decision-making stage, analyzing construction costs helps determine the project's feasibility and economic benefits. However, current technology relies on manual cost analysis, which often requires experienced personnel to provide accurate results. Given the numerous aspects involved in construction cost management, this process is time-consuming and inefficient. Summary of the Invention
[0003] Therefore, the purpose of this application is to provide a construction project cost analysis method and system based on big data analysis, which can overcome the shortcomings of the existing technology.
[0004] To achieve the above objectives, the technical solution adopted in this application is as follows:
[0005] The first aspect of this application provides a method for analyzing construction project costs based on big data analysis, including:
[0006] Obtain an initial construction cost analysis model trained based on big data engineering cost analysis samples;
[0007] Obtain engineering cost analysis training samples of the same engineering type as the construction project to be analyzed; the number of samples in the engineering cost analysis training samples is less than that in the big data engineering cost analysis samples; the engineering cost analysis training samples include engineering cost samples and engineering cost analysis result samples;
[0008] Based on the engineering cost analysis training samples and the initial building engineering cost analysis model, transfer learning training is performed to obtain the target building engineering cost analysis model;
[0009] The cost data of the construction project to be analyzed is input into the cost analysis model of the target construction project to obtain the cost analysis results of the target construction project.
[0010] Compared with traditional technologies, the beneficial effects of this application are:
[0011] The construction cost analysis method based on big data analysis of this application obtains construction cost analysis training samples of the same type as the construction project to be analyzed. Then, it uses construction cost analysis training samples with a smaller sample size than the big data construction cost analysis samples to perform transfer learning training on the already trained initial construction cost analysis model, thereby obtaining a target construction cost analysis model corresponding to the construction cost analysis training samples. Then, the construction cost data of the construction project to be analyzed is input into the target construction cost analysis model, which can efficiently and accurately obtain the target construction cost analysis results of the corresponding construction cost data.
[0012] As one implementation method, the step of obtaining an initial construction project cost analysis model trained based on big data engineering cost analysis samples includes:
[0013] Obtain big data engineering cost analysis samples for multiple engineering types; the big data engineering cost analysis samples include big data engineering cost samples and big data cost analysis result samples;
[0014] The deep learning network model is trained using the big data engineering cost analysis samples to obtain the initial construction engineering cost analysis model.
[0015] In this embodiment, a deep learning network model is trained using big data engineering cost analysis samples from multiple engineering types to obtain a comprehensive initial construction engineering cost analysis model.
[0016] As one implementation method, the step of training a deep learning network model based on the big data engineering cost analysis samples to obtain the initial construction engineering cost analysis model includes:
[0017] Obtain several deep learning network models;
[0018] Based on the first preset number of training iterations, the deep learning network models are trained using the big data engineering cost analysis samples to obtain several first cost analysis models.
[0019] Based on the accuracy of the aforementioned first cost analysis models, several candidate cost analysis models are obtained; the number of candidate cost analysis models is less than the number of first cost analysis models.
[0020] Based on the second preset number of training iterations, the candidate cost analysis models are trained using the big data engineering cost analysis samples to obtain several second cost analysis models; wherein, the second preset number of training iterations is greater than the first preset number of training iterations;
[0021] The second cost analysis model with the highest accuracy was selected as the initial construction project cost analysis model.
[0022] In this embodiment, several candidate cost analysis models are obtained from several first cost analysis models trained according to a first preset number of training times, based on their accuracy. This yields several candidate cost analysis models with high training efficiency, which can be used for training according to a second preset number of training times. This approach can reduce the workload of model training while still obtaining the initial construction cost analysis model with the best training effect.
[0023] As one implementation method, the step of obtaining a plurality of candidate cost analysis models based on the accuracy of the plurality of first cost analysis models includes:
[0024] The accuracy of the aforementioned first cost analysis models is obtained, and accuracy data is acquired.
[0025] Based on the accuracy data, obtain the accuracy threshold;
[0026] Several first cost analysis models whose accuracy is greater than the accuracy threshold are identified as candidate cost analysis models.
[0027] In this embodiment, based on the accuracy of several first cost analysis models, an accuracy threshold can be accurately obtained to screen out candidate cost analysis models whose accuracy is greater than the accuracy threshold.
[0028] As one implementation method, the step of performing transfer learning training based on the engineering cost analysis training samples and the initial building engineering cost analysis model to obtain the target building engineering cost analysis model includes:
[0029] The number of lightweight model parameters is obtained based on the equipment parameters of the target model's operating carrier device and the pre-defined correspondence between the equipment parameters and the number of lightweight model parameters.
[0030] The initial construction cost analysis model is lightweighted based on the number of parameters in the lightweight model to obtain a lightweight construction cost analysis model.
[0031] The lightweight building cost analysis model is trained by transfer learning based on the engineering cost analysis training samples to obtain the target building cost analysis model running on the target model's operating carrier equipment.
[0032] In this embodiment, the corresponding lightweight model parameters are obtained based on the equipment parameters of the target model running carrier device. The initial construction cost analysis model is then lightweighted based on the lightweight model parameters to obtain a lightweight construction cost analysis model suitable for the target model running carrier device. This lightweight model is then used for transfer learning training to obtain a target construction cost analysis model suitable for running on the target model running carrier device.
[0033] A second aspect of this application provides a construction project cost analysis system based on big data analysis, comprising:
[0034] The initial construction cost analysis model acquisition module is used to acquire an initial construction cost analysis model trained based on big data construction cost analysis samples.
[0035] The engineering cost analysis training sample acquisition module is used to acquire engineering cost analysis training samples of the same type as the construction project to be analyzed; the number of samples in the engineering cost analysis training samples is less than that in the big data engineering cost analysis samples; the engineering cost analysis training samples include engineering cost samples and engineering cost analysis result samples;
[0036] The transfer learning module is used to perform transfer learning training based on the engineering cost analysis training samples and the initial building engineering cost analysis model to obtain the target building engineering cost analysis model.
[0037] The analysis module is used to input the cost data of the construction project to be analyzed into the target construction project cost analysis model to obtain the cost analysis results of the target construction project.
[0038] Compared with traditional technologies, the beneficial effects of this application are:
[0039] The construction cost analysis system based on big data analysis of this application obtains construction cost analysis training samples of the same type as the construction project to be analyzed. Then, it uses construction cost analysis training samples with a smaller sample size than the big data construction cost analysis samples to perform transfer learning training on the already trained initial construction cost analysis model, thereby obtaining a target construction cost analysis model corresponding to the construction cost analysis training samples. Then, the construction cost data of the construction project to be analyzed is input into the target construction cost analysis model, which can efficiently and accurately obtain the target construction cost analysis results of the corresponding construction cost data.
[0040] As one implementation method, the initial construction cost analysis model acquisition module includes:
[0041] The big data engineering cost analysis sample acquisition module is used to acquire big data engineering cost analysis samples for multiple engineering types; the big data engineering cost analysis samples include big data engineering cost samples and big data cost analysis result samples;
[0042] The training module is used to perform deep learning training on the deep learning network model based on the big data engineering cost analysis samples to obtain the initial construction engineering cost analysis model.
[0043] In this embodiment, a deep learning network model is trained using big data engineering cost analysis samples from multiple engineering types to obtain a comprehensive initial construction engineering cost analysis model.
[0044] As one implementation method, the training and learning module includes:
[0045] The network model acquisition submodule is used to acquire several deep learning network models;
[0046] The first training submodule is used to train each of the deep learning network models using the big data engineering cost analysis samples according to the first preset training number, so as to obtain a number of first cost analysis models.
[0047] The candidate cost analysis model acquisition submodule is used to acquire a number of candidate cost analysis models based on the accuracy of the number of first cost analysis models; the number of candidate cost analysis models is less than the number of first cost analysis models.
[0048] The second training submodule is used to train each of the candidate cost analysis models using the big data engineering cost analysis samples according to the second preset number of training iterations, so as to obtain a number of second cost analysis models; wherein, the second preset number of training iterations is greater than the first preset number of training iterations.
[0049] The initial construction cost analysis model acquisition submodule is used to determine the second cost analysis model with the highest accuracy as the initial construction cost analysis model.
[0050] In this embodiment, several candidate cost analysis models are obtained from several first cost analysis models trained according to a first preset number of training times, based on their accuracy. This yields several candidate cost analysis models with high training efficiency, which can be used for training according to a second preset number of training times. This approach can reduce the workload of model training while still obtaining the initial construction cost analysis model with the best training effect.
[0051] As one implementation method, the candidate analysis model acquisition submodule includes:
[0052] The accuracy data acquisition submodule is used to acquire the accuracy of the plurality of first cost analysis models and obtain accuracy data;
[0053] An accuracy threshold acquisition submodule is used to acquire an accuracy threshold based on the accuracy data.
[0054] The candidate cost analysis model acquisition submodule is used to identify several first cost analysis models whose accuracy is greater than the accuracy threshold as candidate cost analysis models.
[0055] In this embodiment, based on the accuracy of several first cost analysis models, an accuracy threshold can be accurately obtained to screen out candidate cost analysis models whose accuracy is greater than the accuracy threshold.
[0056] As one implementation, the transfer learning module includes:
[0057] The lightweight model parameter acquisition submodule is used to obtain the lightweight model parameter quantity based on the device parameters of the target model running carrier device and the preset correspondence between the device parameters and the lightweight model parameter quantity.
[0058] The lightweight processing submodule is used to perform lightweight processing on the initial construction cost analysis model according to the number of lightweight model parameters to obtain a lightweight construction cost analysis model.
[0059] The transfer learning submodule is used to perform transfer learning training on the lightweight building engineering cost analysis model based on the engineering cost analysis training samples, so as to obtain the target building engineering cost analysis model running on the target model running carrier device.
[0060] In this embodiment, the corresponding lightweight model parameters are obtained based on the equipment parameters of the target model running carrier device. The initial construction cost analysis model is then lightweighted based on the lightweight model parameters to obtain a lightweight construction cost analysis model suitable for the target model running carrier device. This lightweight model is then used for transfer learning training to obtain a target construction cost analysis model suitable for running on the target model running carrier device.
[0061] To better understand and implement this application, the following detailed description is provided in conjunction with the accompanying drawings. Attached Figure Description
[0062] Figure 1 This is a flowchart illustrating a construction project cost analysis method based on big data analysis according to an embodiment of this application;
[0063] Figure 2 This is a module connection diagram of a construction project cost analysis system based on big data analysis according to an embodiment of this application;
[0064] 100. Construction project cost analysis system; 101. Initial construction project cost analysis model acquisition module; 102. Construction project cost analysis training sample acquisition module; 103. Transfer learning module; 104. Analysis module. Detailed Implementation
[0065] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.
[0066] It should be understood that the described embodiments are merely some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of the embodiments of this application.
[0067] In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. In the description of this application, it should be understood that the terms "first," "second," "third," etc., are used only to distinguish similar objects and are not necessarily used to describe a specific order or sequence, nor should they be construed as indicating or implying relative importance. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances. The singular forms "a," "the," and "the" used in this application and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. The word "if" as used herein can be interpreted as "when," "when," or "in response to determination."
[0068] Furthermore, in the description of this application, unless otherwise stated, "multiple" means two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. The character " / " generally indicates that the preceding and following related objects have an "or" relationship.
[0069] Please see Figure 1 This is a flowchart of the construction project cost analysis method based on big data analysis according to the first embodiment of this application. The method includes:
[0070] S1: Obtain an initial construction cost analysis model trained based on big data engineering cost analysis samples;
[0071] The big data engineering cost analysis samples include engineering cost analysis samples of various engineering types, such as factories, theaters, hotels, shops, schools, hospitals, office buildings, and residences.
[0072] S2: Obtain engineering cost analysis training samples of the same engineering type as the construction project to be analyzed; the number of samples in the engineering cost analysis training samples is less than that in the big data engineering cost analysis samples; the engineering cost analysis training samples include engineering cost samples and engineering cost analysis result samples.
[0073] S3: Based on the engineering cost analysis training samples and the initial building engineering cost analysis model, perform transfer learning training to obtain the target building engineering cost analysis model;
[0074] S4: Input the cost data of the construction project to be analyzed into the target construction project cost analysis model to obtain the target construction project cost analysis results.
[0075] In this context, the cost analysis results will vary depending on the type of construction project being analyzed. While an initial construction cost analysis model can analyze cost data for multiple project types, its accuracy is lower than a model specifically designed for that particular project type. This application, however, utilizes transfer learning training. With a small number of cost analysis training samples, a target construction cost analysis model with higher accuracy for the specific project type being analyzed can be trained. Furthermore, the smaller number of training samples required for transfer learning reduces training time and improves efficiency, thereby achieving the goal of efficiently and accurately obtaining the target construction cost analysis results from the analyzed project cost data.
[0076] The cost analysis results of a target construction project are obtained through data analysis of the cost data of the project being analyzed. Data analysis can be divided into longitudinal analysis and cross-sectional analysis. Longitudinal analysis mainly studies the cost changes of a project during construction, the proportion of each unit project in the total project cost, the usage and distribution of various major materials, and various factors affecting the cost. Cross-sectional analysis mainly studies the differences in cost among similar projects and the reasons for these differences, identifying the common cost patterns reflected in similar projects.
[0077] Based on the scope of the analysis object, engineering cost analysis can be divided into overall level analysis, component analysis, influencing factors and risk analysis, and change analysis.
[0078] (1) Overall level analysis: Analysis of the degree of fit between information reflecting the project cost status (work volume, resource consumption, investment) and the construction scale.
[0079] (2) Composition analysis: refers to the proportional relationship of various components in the project cost. Taking the total cost analysis of a construction project as an example, it includes the proportion of building construction costs, equipment costs, installation costs, and deferred expenses to the total cost.
[0080] (3) Influencing Factors and Risk Analysis: This refers to the evaluation and analysis of the main influencing factors, characteristics, and potential risks in the process of project cost formation. The influencing factors of project cost include not only "quantity" and "price," but also project overview, construction conditions, project characteristics, etc.
[0081] (4) Variation Analysis: This refers to analyzing the object of analysis from a changing perspective compared to a comparable target. For example, the longitudinal comparison of technical and economic indicators at different stages of the same project aims to analyze the control of project costs at different stages and to analyze overspending or cost savings. Similarly, the longitudinal comparison of technical and economic indicators at different periods of the same type of project aims to analyze the changing trends of project cost technical and economic indicators at different times.
[0082] Compared with traditional technologies, the beneficial effects of this application are:
[0083] The construction cost analysis method based on big data analysis of this application obtains construction cost analysis training samples of the same type as the construction project to be analyzed. Then, it uses construction cost analysis training samples with a smaller sample size than the big data construction cost analysis samples to perform transfer learning training on the already trained initial construction cost analysis model, thereby obtaining a target construction cost analysis model corresponding to the construction cost analysis training samples. Then, the construction cost data of the construction project to be analyzed is input into the target construction cost analysis model, which can efficiently and accurately obtain the target construction cost analysis results of the corresponding construction cost data.
[0084] In a feasible embodiment, step S1: obtaining an initial construction cost analysis model trained based on big data construction cost analysis samples, includes:
[0085] S11: Obtain big data engineering cost analysis samples for multiple engineering types; the big data engineering cost analysis samples include big data engineering cost samples and big data cost analysis result samples;
[0086] S12: Based on the big data engineering cost analysis samples, perform deep learning training on the deep learning network model to obtain the initial construction engineering cost analysis model.
[0087] Specifically, the big data engineering cost sample is used as input, and the corresponding big data cost analysis result sample is used as output to train the deep learning network model to obtain the initial construction engineering cost analysis model.
[0088] In this embodiment, a deep learning network model is trained using big data engineering cost analysis samples from multiple engineering types to obtain a comprehensive initial building engineering cost analysis model.
[0089] In a feasible embodiment, step S12: training a deep learning network model based on the big data engineering cost analysis samples to obtain the initial construction engineering cost analysis model includes:
[0090] S121: Obtain several deep learning network models;
[0091] S122: Based on the first preset number of training iterations, the deep learning network models are trained using the big data engineering cost analysis samples to obtain several first cost analysis models;
[0092] Because the training of each deep learning network model is random during the training process, several different first cost analysis models will be obtained under the same number of training iterations.
[0093] S123: Based on the accuracy of the plurality of first cost analysis models, obtain a plurality of candidate cost analysis models; the number of candidate cost analysis models is less than the number of first cost analysis models;
[0094] The higher the accuracy of the first cost analysis model, the better the training effect of the corresponding first cost analysis model, and the better it can be used as a candidate cost analysis model for further training.
[0095] S124: Based on the second preset number of training iterations, the candidate cost analysis models are trained using the big data engineering cost analysis samples to obtain several second cost analysis models; wherein, the second preset number of training iterations is greater than the first preset number of training iterations;
[0096] Since the second preset number of training iterations is greater than the first preset number of training iterations, training a single model using the second preset number of iterations will require more time and resources than training using the first preset number of iterations. Training only the candidate cost analysis model reduces the number of network models that need to be trained, thereby reducing the overall training time and required resources.
[0097] S125: The second cost analysis model with the highest accuracy is determined as the initial construction project cost analysis model.
[0098] In this embodiment, several candidate cost analysis models are obtained from several first cost analysis models trained according to a first preset number of training times, based on their accuracy. This yields several candidate cost analysis models with high training efficiency, which can be used for training according to a second preset number of training times. This approach can reduce the workload of model training while still obtaining the initial construction cost analysis model with the best training effect.
[0099] In a feasible embodiment, step S123: obtaining a plurality of candidate cost analysis models based on the accuracy of the plurality of first cost analysis models, includes:
[0100] S1231: Obtain the accuracy of the plurality of first cost analysis models and obtain accuracy data;
[0101] S1232: Obtain the accuracy threshold based on the accuracy data;
[0102] The accuracy threshold can be the average value of the accuracy data.
[0103] S1233: Several first cost analysis models with accuracy greater than the accuracy threshold are identified as candidate cost analysis models.
[0104] In this embodiment, based on the accuracy of several first cost analysis models, an accuracy threshold can be accurately obtained to screen out candidate cost analysis models whose accuracy is greater than the accuracy threshold.
[0105] In a feasible embodiment, step S3: performing transfer learning training based on the engineering cost analysis training samples and the initial construction engineering cost analysis model to obtain the target construction engineering cost analysis model includes:
[0106] S31: Based on the equipment parameters of the target model's operating carrier equipment and the pre-defined correspondence between the equipment parameters and the lightweight model's parameter quantities, obtain the lightweight model's parameter quantities;
[0107] The more parameters a network model has, the higher its accuracy. However, the more parameters a network model has, the higher its requirements for device parameters. The correspondence between device parameters and the number of lightweight model parameters refers to the maximum number of parameters that the model's runtime equipment can support for the smooth operation of the network model. Based on the correspondence between device parameters and the number of lightweight model parameters, we can obtain the maximum number of parameters that the target model's runtime equipment can support for the smooth operation of the lightweight model, i.e., the number of lightweight model parameters.
[0108] S32: The initial construction cost analysis model is lightweighted according to the number of lightweight model parameters to obtain a lightweight construction cost analysis model.
[0109] Lightweighting refers to reducing the computational and storage requirements of a network model through techniques such as model pruning and quantization, enabling the network model to run smoothly on low-power devices while minimizing its performance impact. However, the more model parameters are reduced through lightweighting, the greater the impact on network model performance. Therefore, lightweighting the initial construction cost analysis model based on the corresponding number of lightweight model parameters for the device parameters can balance the accuracy of the lightweight construction cost analysis model with the computational requirements, thus obtaining a lightweight construction cost analysis model suitable for running on the target model's operating platform.
[0110] S33: Based on the engineering cost analysis training samples, the lightweight building engineering cost analysis model is trained by transfer learning to obtain the target building engineering cost analysis model running on the target model's operating carrier equipment.
[0111] Among them, the target construction cost analysis model obtained through transfer learning can accurately analyze the construction cost data of the construction project to be analyzed for the project type, so as to efficiently and accurately obtain the target construction cost analysis results of the construction cost data to be analyzed.
[0112] In this embodiment, the corresponding lightweight model parameters are obtained based on the equipment parameters of the target model running carrier device. The initial construction cost analysis model is then lightweighted based on the lightweight model parameters to obtain a lightweight construction cost analysis model suitable for the target model running carrier device. This lightweight model is then used for transfer learning training to obtain a target construction cost analysis model suitable for running on the target model running carrier device.
[0113] Please see Figure 2 The second aspect of this application provides a construction project cost analysis system based on big data analysis, comprising:
[0114] The initial construction cost analysis model acquisition module is used to acquire an initial construction cost analysis model trained based on big data construction cost analysis samples.
[0115] The engineering cost analysis training sample acquisition module is used to acquire engineering cost analysis training samples of the same type as the construction project to be analyzed; the number of samples in the engineering cost analysis training samples is less than that in the big data engineering cost analysis samples; the engineering cost analysis training samples include engineering cost samples and engineering cost analysis result samples;
[0116] The transfer learning module is used to perform transfer learning training based on the engineering cost analysis training samples and the initial building engineering cost analysis model to obtain the target building engineering cost analysis model.
[0117] The analysis module is used to input the cost data of the construction project to be analyzed into the target construction project cost analysis model to obtain the cost analysis results of the target construction project.
[0118] Compared with traditional technologies, the beneficial effects of this application are:
[0119] The construction cost analysis system based on big data analysis of this application obtains construction cost analysis training samples of the same type as the construction project to be analyzed. Then, it uses construction cost analysis training samples with a smaller sample size than the big data construction cost analysis samples to perform transfer learning training on the already trained initial construction cost analysis model, thereby obtaining a target construction cost analysis model corresponding to the construction cost analysis training samples. Then, the construction cost data of the construction project to be analyzed is input into the target construction cost analysis model, which can efficiently and accurately obtain the target construction cost analysis results of the corresponding construction cost data.
[0120] In one feasible embodiment, the initial construction cost analysis model acquisition module includes:
[0121] The big data engineering cost analysis sample acquisition module is used to acquire big data engineering cost analysis samples for multiple engineering types; the big data engineering cost analysis samples include big data engineering cost samples and big data cost analysis result samples;
[0122] The training module is used to perform deep learning training on the deep learning network model based on the big data engineering cost analysis samples to obtain the initial construction engineering cost analysis model.
[0123] In this embodiment, a deep learning network model is trained using big data engineering cost analysis samples from multiple engineering types to obtain a comprehensive initial building engineering cost analysis model.
[0124] In one feasible embodiment, the training and learning module includes:
[0125] The network model acquisition submodule is used to acquire several deep learning network models;
[0126] The first training submodule is used to train each of the deep learning network models using the big data engineering cost analysis samples according to the first preset training number, so as to obtain a number of first cost analysis models.
[0127] The candidate cost analysis model acquisition submodule is used to acquire a number of candidate cost analysis models based on the accuracy of the number of first cost analysis models; the number of candidate cost analysis models is less than the number of first cost analysis models.
[0128] The second training submodule is used to train each of the candidate cost analysis models using the big data engineering cost analysis samples according to the second preset number of training iterations, so as to obtain a number of second cost analysis models; wherein, the second preset number of training iterations is greater than the first preset number of training iterations.
[0129] The initial construction cost analysis model acquisition submodule is used to determine the second cost analysis model with the highest accuracy as the initial construction cost analysis model.
[0130] In this embodiment, several candidate cost analysis models are obtained from several first cost analysis models trained according to a first preset number of training times, based on their accuracy. This yields several candidate cost analysis models with high training efficiency, which can be used for training according to a second preset number of training times. This approach can reduce the workload of model training while still obtaining the initial construction cost analysis model with the best training effect.
[0131] In one feasible embodiment, the candidate analysis model acquisition submodule includes:
[0132] The accuracy data acquisition submodule is used to acquire the accuracy of the plurality of first cost analysis models and obtain accuracy data;
[0133] An accuracy threshold acquisition submodule is used to acquire an accuracy threshold based on the accuracy data.
[0134] The candidate cost analysis model acquisition submodule is used to identify several first cost analysis models whose accuracy is greater than the accuracy threshold as candidate cost analysis models.
[0135] In this embodiment, based on the accuracy of several first cost analysis models, an accuracy threshold can be accurately obtained to screen out candidate cost analysis models whose accuracy is greater than the accuracy threshold.
[0136] In one feasible embodiment, the transfer learning module includes:
[0137] The lightweight model parameter acquisition submodule is used to obtain the lightweight model parameter quantity based on the device parameters of the target model running carrier device and the preset correspondence between the device parameters and the lightweight model parameter quantity.
[0138] The lightweight processing submodule is used to perform lightweight processing on the initial construction cost analysis model according to the number of lightweight model parameters to obtain a lightweight construction cost analysis model.
[0139] The transfer learning submodule is used to perform transfer learning training on the lightweight building engineering cost analysis model based on the engineering cost analysis training samples, so as to obtain the target building engineering cost analysis model running on the target model running carrier device.
[0140] In this embodiment, the corresponding lightweight model parameters are obtained based on the equipment parameters of the target model running carrier device. The initial construction cost analysis model is then lightweighted based on the lightweight model parameters to obtain a lightweight construction cost analysis model suitable for the target model running carrier device. This lightweight model is then used for transfer learning training to obtain a target construction cost analysis model suitable for running on the target model running carrier device.
[0141] It should be noted that the construction cost analysis system based on big data analysis provided in the second aspect of this application is only illustrated by the above-described division of functional modules when executing the construction cost analysis method based on big data analysis. In practical applications, the above functions can be assigned to different functional modules as needed, that is, the internal structure of the equipment can be divided into different functional modules to complete all or part of the functions described above. Furthermore, the construction cost analysis system based on big data analysis provided in the second aspect of this application and the construction cost analysis method based on big data analysis in the first aspect of this application belong to the same concept, and their implementation process is detailed in the method embodiments, which will not be repeated here.
[0142] The device embodiments described above are merely illustrative. The components described as separate parts may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without any inventive effort.
[0143] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.
[0144] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 The computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function selected in one or more boxes.
[0145] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function selected in one or more boxes.
[0146] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.
[0147] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.
[0148] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.
[0149] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0150] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.
Claims
1. A method for analyzing construction project costs based on big data analysis, characterized in that, include: Obtain an initial construction cost analysis model trained based on big data engineering cost analysis samples; Obtain engineering cost analysis training samples of the same engineering type as the construction project to be analyzed; the number of samples in the engineering cost analysis training samples is less than that in the big data engineering cost analysis samples; the engineering cost analysis training samples include engineering cost samples and engineering cost analysis result samples; The target building cost analysis model is obtained by performing transfer learning training based on the engineering cost analysis training samples and the initial building engineering cost analysis model. The cost data of the construction project to be analyzed is input into the cost analysis model of the target construction project to obtain the cost analysis results of the target construction project.
2. The construction project cost analysis method based on big data analysis according to claim 1, characterized in that, The steps for obtaining the initial construction cost analysis model trained based on big data construction cost analysis samples include: Obtain big data engineering cost analysis samples for multiple engineering types; the big data engineering cost analysis samples include big data engineering cost samples and big data cost analysis result samples; The deep learning network model is trained using the big data engineering cost analysis samples to obtain the initial construction engineering cost analysis model.
3. The construction project cost analysis method based on big data analysis according to claim 1, characterized in that, The step of training a deep learning network model based on the big data engineering cost analysis samples to obtain the initial construction engineering cost analysis model includes: Obtain several deep learning network models; Based on the first preset number of training iterations, the deep learning network models are trained using the big data engineering cost analysis samples to obtain several first cost analysis models. Based on the accuracy of the aforementioned first cost analysis models, several candidate cost analysis models are obtained; the number of candidate cost analysis models is less than the number of first cost analysis models. Based on the second preset number of training iterations, the candidate cost analysis models are trained using the big data engineering cost analysis samples to obtain several second cost analysis models; wherein, the second preset number of training iterations is greater than the first preset number of training iterations; The second cost analysis model with the highest accuracy was selected as the initial construction project cost analysis model.
4. The construction project cost analysis method based on big data analysis according to claim 3, characterized in that, The step of obtaining several candidate cost analysis models based on the accuracy of the several first cost analysis models includes: The accuracy of the aforementioned first cost analysis models is obtained, and accuracy data is acquired. Based on the accuracy data, obtain the accuracy threshold; Several first cost analysis models whose accuracy is greater than the accuracy threshold are identified as candidate cost analysis models.
5. The construction project cost analysis method based on big data analysis according to claim 1, characterized in that, The step of performing transfer learning training based on the engineering cost analysis training samples and the initial building engineering cost analysis model to obtain the target building engineering cost analysis model includes: The number of lightweight model parameters is obtained based on the equipment parameters of the target model's operating carrier device and the pre-defined correspondence between the equipment parameters and the number of lightweight model parameters. The initial construction cost analysis model is lightweighted based on the number of parameters in the lightweight model to obtain a lightweight construction cost analysis model. The lightweight building cost analysis model is trained by transfer learning based on the engineering cost analysis training samples to obtain the target building cost analysis model running on the target model's operating carrier equipment.
6. A construction project cost analysis system based on big data analysis, characterized in that, include: The initial construction cost analysis model acquisition module is used to acquire an initial construction cost analysis model trained based on big data construction cost analysis samples. The engineering cost analysis training sample acquisition module is used to acquire engineering cost analysis training samples of the same type as the construction project to be analyzed; the number of samples in the engineering cost analysis training samples is less than that in the big data engineering cost analysis samples; the engineering cost analysis training samples include engineering cost samples and engineering cost analysis result samples; The transfer learning module is used to perform transfer learning training based on the engineering cost analysis training samples and the initial building engineering cost analysis model to obtain the target building engineering cost analysis model. The analysis module is used to input the cost data of the construction project to be analyzed into the target construction project cost analysis model to obtain the cost analysis results of the target construction project.
7. The construction project cost analysis system based on big data analysis according to claim 6, characterized in that, The initial construction cost analysis model acquisition module includes: The big data engineering cost analysis sample acquisition module is used to acquire big data engineering cost analysis samples for multiple engineering types; the big data engineering cost analysis samples include big data engineering cost samples and big data cost analysis result samples; The training module is used to perform deep learning training on the deep learning network model based on the big data engineering cost analysis samples to obtain the initial construction engineering cost analysis model.
8. The construction project cost analysis system based on big data analysis according to claim 5, characterized in that, The training and learning module includes: The network model acquisition submodule is used to acquire several deep learning network models; The first training submodule is used to train each of the deep learning network models using the big data engineering cost analysis samples according to the first preset training number, so as to obtain a number of first cost analysis models. The candidate cost analysis model acquisition submodule is used to acquire a number of candidate cost analysis models based on the accuracy of the number of first cost analysis models; the number of candidate cost analysis models is less than the number of first cost analysis models. The second training submodule is used to train each of the candidate cost analysis models using the big data engineering cost analysis samples according to the second preset number of training iterations, so as to obtain a number of second cost analysis models; wherein, the second preset number of training iterations is greater than the first preset number of training iterations. The initial construction cost analysis model acquisition submodule is used to determine the second cost analysis model with the highest accuracy as the initial construction cost analysis model.
9. The construction project cost analysis system based on big data analysis according to claim 8, characterized in that, The candidate analysis model acquisition submodule includes: The accuracy data acquisition submodule is used to acquire the accuracy of the plurality of first cost analysis models and obtain accuracy data; An accuracy threshold acquisition submodule is used to acquire an accuracy threshold based on the accuracy data. The candidate cost analysis model acquisition submodule is used to identify several first cost analysis models whose accuracy is greater than the accuracy threshold as candidate cost analysis models.
10. The construction project cost analysis system based on big data analysis according to claim 6, characterized in that, The transfer learning module includes: The lightweight model parameter acquisition submodule is used to obtain the lightweight model parameter quantity based on the device parameters of the target model running carrier device and the preset correspondence between the device parameters and the lightweight model parameter quantity. The lightweight processing submodule is used to perform lightweight processing on the initial construction cost analysis model according to the number of lightweight model parameters to obtain a lightweight construction cost analysis model. The transfer learning submodule is used to perform transfer learning training on the lightweight building engineering cost analysis model based on the engineering cost analysis training samples, so as to obtain the target building engineering cost analysis model running on the target model running carrier device.