Construction site carbon emission monitoring method, device and equipment based on multi-source data fusion

Abstract

The application relates to the engineering technical field and discloses a construction site carbon emission monitoring method, device and equipment based on multi-source data fusion, which comprises the following steps: adopting a text feature extraction network to respectively extract features of a bill of quantities of a current construction project, a construction organization design file of the current construction project and a construction schedule of the current construction project, and respectively obtaining planned material consumption of the current construction project, mechanical parameters of the current construction project and a planned construction period of the current construction project; adopting an attention network to fuse the planned material consumption of the current construction project, the mechanical parameters of the current construction project and the planned construction period of the current construction project, and obtaining a fusion feature vector of the current construction project; inputting the current fusion feature vector of the current construction project into a large language model after training; and generating a predicted carbon emission of the current construction project at a current construction site through the large language model after training. The application can improve the timeliness of carbon emission evaluation.

Description

Technical Field

[0001] This application relates to the field of engineering technology, and in particular to a method, apparatus and equipment for monitoring carbon emissions at construction sites based on multi-source data fusion. Background Technology

[0002] During the construction process, various construction machinery and equipment consume fuel, electricity and other energy, and energy consumption inevitably leads to a large amount of carbon emissions, making the construction process one of the important sources of carbon emissions.

[0003] However, most existing technologies estimate carbon emissions after completion, making it difficult to obtain pre-construction carbon emission forecasts in a timely manner. This lack of data support for optimizing and adjusting construction plans increases the risk of exceeding carbon emission limits during construction. Therefore, obtaining the predicted carbon emissions of a current construction project at the current site before construction begins is a pressing technical problem that needs to be solved. Summary of the Invention

[0004] This application provides a method, apparatus, and equipment for monitoring carbon emissions at construction sites based on multi-source data fusion, in order to solve the aforementioned technical problem of how to obtain the predicted carbon emissions of a current construction project at the current construction site.

[0005] In a first aspect, embodiments of this application provide a construction site carbon emission monitoring method based on multi-source data fusion, applied to electronic devices, the construction site carbon emission monitoring method comprising: Acquire multi-source data for the preset construction project. The multi-source data for the preset construction project includes the bill of quantities, the construction organization design document, and the construction schedule of the preset construction project. A text feature extraction network is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the pre-defined construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the pre-defined construction project. An attention network is used to fuse the planned material usage, mechanical parameters, and planned construction period of a pre-defined construction project to obtain a fused feature vector. The fused feature vector and the actual carbon emissions of the pre-defined construction project at the pre-defined construction site are combined to form a sample. Different samples are combined to form a dataset. The pre-defined large language model is trained using the dataset, and the loss value of the large language model on the dataset is obtained. When the loss value of the large language model on the dataset no longer decreases, the training of the large language model is stopped, and the trained large language model is saved. Send a data acquisition request for the current construction project to the database, and receive multi-source data of the current construction project returned by the database based on the data acquisition request. The multi-source data of the current construction project includes the bill of quantities, the construction organization design documents, and the construction schedule of the current construction project. A text feature extraction network is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the current construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the current construction project. An attention network is then used to fuse these three data to obtain a fused feature vector. This fused feature vector is then input into a trained large language model, which is used to generate the predicted carbon emissions of the current construction project at the current construction site.

[0006] In one possible implementation of the first aspect, multi-source data of a pre-defined construction project is obtained. This multi-source data includes a bill of quantities for the pre-defined construction project, a construction organization design document for the pre-defined construction project, and a construction schedule for the pre-defined construction project, including: The system sends connection requests to the file storage system via a communication network and receives response messages from the file storage system based on the connection requests. The communication network includes Ethernet and WIFI networks. When the response message is an authentication success message, multi-source data of the preset construction project is retrieved from the file storage system. The multi-source data of the preset construction project includes the bill of quantities of the preset construction project, the construction organization design document of the preset construction project, and the construction schedule of the preset construction project.

[0007] In one possible implementation of the first aspect, an attention network is employed to fuse the planned material usage, mechanical parameters, and planned construction period of a pre-defined construction project, resulting in a fused feature vector. This fused feature vector, along with the actual carbon emissions of the project at the pre-defined construction site, forms a sample. Different samples are combined into a dataset, which is then used to train a pre-defined large language model. The loss value of the large language model on the dataset is obtained. When the loss value of the large language model on the dataset no longer decreases, training of the large language model is stopped, and the trained large language model is saved, including: Input the planned material usage, mechanical parameters, and planned construction period of the pre-defined construction project into the attention network; An attention network is used to fuse the planned material usage, mechanical parameters, and planned construction period of a pre-designed construction project to obtain a fused feature vector of the pre-designed construction project. The fused feature vector of the pre-designed construction project and the actual carbon emissions of the pre-designed construction project at the pre-designed construction site are then combined to form a sample. Different samples are combined into a dataset, and the dataset is used to train a pre-defined large language model. The loss value of the large language model on the dataset is obtained through the mean squared error loss function. The goal is to reduce the loss value. When the loss value of the large language model on the dataset no longer decreases, the training of the large language model is stopped, and the trained large language model is saved.

[0008] In one possible implementation of the first aspect, a text feature extraction network is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the current construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned duration of the current construction project. An attention network is then used to fuse these features to obtain a fused feature vector. This fused feature vector is input into a trained large language model. After generating the predicted carbon emissions of the current construction project at the current construction site using the trained large language model, a site carbon emission monitoring method is implemented, including: When the predicted carbon emissions of the current construction project at the current site are not less than the preset value, emission reduction suggestions for the current construction project will be displayed. These suggestions include reducing the operating time of the target machinery and adjusting the construction sequence. If the predicted carbon emissions of the current construction project at the current site are less than the preset value, a message indicating that the current construction project has been approved will be displayed.

[0009] In one possible implementation of the first aspect, after displaying a pass message for the current construction project when the predicted carbon emissions of the current construction project at the current site are less than a preset value, the site carbon emission monitoring method includes: Obtain monitoring parameters for the current construction project, including dust concentration, noise levels, and energy consumption. When the dust concentration exceeds the preset concentration, a start command is sent to the fog cannon; when the dust concentration drops back to the qualified range, a shut-off command is sent to the fog cannon. When the noise level exceeds the preset noise limit, noise reduction suggestions are output, including adjusting the process or replacing with low-noise equipment. When the noise level does not exceed the preset noise limit, a noise compliance message is output. When energy consumption exceeds the preset consumption, optimization suggestions are output, including shortening the single running time of the process equipment and delaying the start-up time of the process equipment to the off-peak electricity price period.

[0010] In one possible implementation of the first aspect, monitoring parameters of the current construction project are obtained, including dust concentration, noise levels (decibels), and energy consumption, among others: Obtain the monitoring time for the current construction project and determine whether the current time is a monitoring time. If the current time is the monitoring time, obtain the monitoring parameters of the current construction project, including dust concentration, noise decibels, and energy consumption.

[0011] In one possible implementation of the first aspect, the pre-set mechanical parameters of the construction project include the rated power of each type of machinery in the pre-set construction project, the planned quantity of each type of machinery in the pre-set construction project, and the planned operating time of each type of machinery in the pre-set construction project.

[0012] In one possible implementation of the first aspect, the mechanical parameters of the current construction project include the rated power of each type of machinery in the current construction project, the planned quantity of each type of machinery in the current construction project, and the planned operating time of each type of machinery in the current construction project.

[0013] Secondly, embodiments of this application provide a construction site carbon emission monitoring device based on multi-source data fusion, applied to electronic equipment, including: The first acquisition module is used to acquire multi-source data of the preset construction project. The multi-source data of the preset construction project includes the bill of quantities of the preset construction project, the construction organization design document of the preset construction project, and the construction schedule of the preset construction project. The extraction module is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the preset construction project using a text feature extraction network, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the preset construction project. The fusion module uses an attention network to fuse the planned material usage, mechanical parameters, and planned construction period of a pre-defined construction project to obtain a fused feature vector. This fused feature vector and the actual carbon emissions of the pre-defined construction project at the pre-defined construction site are combined to form a sample. Different samples are combined to form a dataset. The dataset is used to train a pre-defined large language model, and the loss value of the large language model on the dataset is obtained. When the loss value of the large language model on the dataset no longer decreases, the training of the large language model is stopped, and the trained large language model is saved. The second acquisition module is used to send a data acquisition request for the current construction project to the database and receive multi-source data of the current construction project returned by the database according to the data acquisition request. The multi-source data of the current construction project includes the bill of quantities of the current construction project, the construction organization design document of the current construction project, and the construction schedule of the current construction project. The generation module uses a text feature extraction network to extract features from the bill of quantities, construction organization design documents, and construction schedule of the current construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period. An attention network is then used to fuse these features to obtain a fused feature vector. This fused feature vector is input into a trained large language model, which then generates the predicted carbon emissions of the current construction project at the current construction site.

[0014] Thirdly, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the construction site carbon emission monitoring method described in the first aspect above.

[0015] The beneficial effects of the embodiments of this application are as follows: Firstly, a text feature extraction network is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the current construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the current construction project. An attention network is then used to fuse these features to obtain a fused feature vector. This fused feature vector is input into a trained large language model, which generates the predicted carbon emissions of the current construction project at the current construction site. This method can quickly generate carbon emission prediction results before construction begins, which shortens the time for obtaining predicted carbon emissions and improves the efficiency of obtaining predicted carbon emissions compared to traditional post-construction statistical methods, thus improving the timeliness of carbon emission assessment. Secondly, since the predicted carbon emissions of the current construction project at the current site are automatically completed by the trained large language model, there is no need to rely on manual on-site surveys and experience-based judgments. The entire prediction process is not affected by factors such as the subjective experience of operators, differences in judgment standards, and human negligence, thus ensuring the objectivity and stability of the predicted carbon emissions and improving the reliability of the predicted carbon emissions. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This application scenario diagram illustrates the construction site carbon emission monitoring method provided in the embodiments of this application. Figure 2 This is a flowchart illustrating the construction site carbon emission monitoring method provided in the embodiments of this application; Figure 3 A flowchart illustrating the implementation of S203 provided in this application embodiment; Figure 4 A schematic block diagram of a construction site carbon emission monitoring device provided in the embodiments of this application; Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.

[0019] The construction site carbon emission monitoring method provided in this application embodiment can be applied to electronic devices, including but not limited to servers, mobile phones, tablets, and laptops. This application embodiment does not impose any restrictions on the specific type of electronic device.

[0020] Please see Figure 1 , Figure 1 The application scenario diagram of the construction site carbon emission monitoring method provided in the embodiments of this application is described in detail below: Electronic devices send connection requests to file storage systems via communication networks and receive response messages from file storage systems in response to the connection requests. Communication networks include Ethernet and Wi-Fi networks. When the response message is an authentication success message, multi-source data of the preset construction project is retrieved from the file storage system. The multi-source data of the preset construction project includes the bill of quantities of the preset construction project, the construction organization design document of the preset construction project, and the construction schedule of the preset construction project.

[0021] In this embodiment, the electronic device obtains multi-source data of the preset construction project from the file storage system, avoiding data omissions or input errors that may be introduced by manual operation, and improving the accuracy of the multi-source data of the preset construction project.

[0022] Please see Figure 2 , Figure 2 This is a flowchart illustrating the construction site carbon emission monitoring method provided in this application embodiment, which can be applied to electronic devices.

[0023] like Figure 2 As shown in the embodiments of this application, the construction site carbon emission monitoring method includes the following steps, which are detailed below: S201, Obtain multi-source data of the preset construction project. The multi-source data of the preset construction project includes the bill of quantities of the preset construction project, the construction organization design document of the preset construction project, and the construction schedule of the preset construction project. The bill of quantities for a pre-designed construction project refers to a detailed list of all the engineering contents and their corresponding quantities in the pre-designed construction project.

[0024] The construction organization design document for a pre-planned construction project refers to a comprehensive technical document concerning the overall construction deployment, construction methods, construction sequence, resource allocation, and site layout of the pre-planned construction project.

[0025] The pre-set construction schedule table refers to the time arrangement and interrelationship of each construction procedure and stage in the pre-set construction project, including the start time, completion time, duration of each procedure, and logical relationship between procedures.

[0026] S202 uses a text feature extraction network to extract features from the bill of quantities, construction organization design documents, and construction schedule of the pre-defined construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the pre-defined construction project. Text feature extraction networks are neural network models used to transform raw text data into numerical feature vectors, enabling computers to understand and process text information. Text feature extraction networks include, but are not limited to, the following: Word embedding model: Maps each word in the text to a dense vector of fixed length, so that words with similar meanings are adjacent to each other in the vector space.

[0027] Convolutional Neural Networks (CNNs): By using a one-dimensional convolutional kernel to perform a sliding window operation on a text sequence, they can effectively capture semantic information of phrases and local patterns. They are suitable for scenarios requiring the extraction of key local features, such as text classification and sentiment analysis.

[0028] S203 employs an attention network to fuse the planned material usage, mechanical parameters, and planned construction period of a pre-defined construction project, resulting in a fused feature vector. This fused feature vector and the actual carbon emissions of the pre-defined construction project at the pre-defined construction site are combined to form a sample. Different samples are combined to form a dataset. The dataset is used to train a pre-defined large language model, and the loss value of the large language model on the dataset is obtained. When the loss value of the large language model on the dataset no longer decreases, training of the large language model is stopped, and the trained large language model is saved. Attention networks, in particular, are neural network structures that enable large language models to dynamically focus on information that is more important to the current task by calculating the relevance weights between elements in the input sequence.

[0029] S204, send a data acquisition request for the current construction project to the database, and receive the multi-source data of the current construction project returned by the database according to the data acquisition request. The multi-source data of the current construction project includes the bill of quantities of the current construction project, the construction organization design document of the current construction project, and the construction schedule of the current construction project. S205 employs a text feature extraction network to extract features from the bill of quantities, construction organization design documents, and construction schedule of the current construction project, respectively, obtaining the planned material usage, mechanical parameters, and planned construction period. An attention network is then used to fuse these features, resulting in a fused feature vector. This fused feature vector is input into a trained large language model, which then generates a predicted carbon emission level for the current construction project at the current construction site.

[0030] Specifically, in the text feature extraction network, features are extracted from the bill of quantities, construction organization design documents, and construction schedule of the current construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the current construction project. An attention network is then used to fuse these three data points to obtain a fused feature vector. This fused feature vector is input into a trained large language model, which then generates the predicted carbon emissions of the current construction project at the current construction site. The site carbon emission monitoring method includes: When the predicted carbon emissions of the current construction project at the current site are not less than the preset value, emission reduction suggestions for the current construction project will be displayed. These suggestions include reducing the operating time of the target machinery and adjusting the construction sequence. If the predicted carbon emissions of the current construction project at the current site are less than the preset value, a message indicating that the current construction project has been approved will be displayed.

[0031] Wherein, after displaying a pass message for the current construction project when the predicted carbon emissions of the current construction project at the current site are less than a preset value, the site carbon emission monitoring method includes: Obtain monitoring parameters for the current construction project, including dust concentration, noise levels, and energy consumption. When the dust concentration exceeds the preset concentration, a start command is sent to the fog cannon; when the dust concentration drops back to the qualified range, a shut-off command is sent to the fog cannon. When the noise level exceeds the preset noise limit, noise reduction suggestions are output, including adjusting the process or replacing with low-noise equipment. When the noise level does not exceed the preset noise limit, a noise compliance message is output. When energy consumption exceeds the preset consumption, optimization suggestions are output, including shortening the single running time of the process equipment and delaying the start-up time of the process equipment to the off-peak electricity price period.

[0032] This includes acquiring monitoring parameters for the current construction project, such as dust concentration, noise levels, and energy consumption. Obtain the monitoring time for the current construction project and determine whether the current time is a monitoring time. If the current time is the monitoring time, obtain the monitoring parameters of the current construction project, including dust concentration, noise decibels, and energy consumption.

[0033] The mechanical parameters of the pre-set construction project include the rated power of each type of machinery in the pre-set construction project, the planned quantity of each type of machinery in the pre-set construction project, and the planned operating time of each type of machinery in the pre-set construction project.

[0034] The mechanical parameters of the current construction project include the rated power of each type of machinery in the current construction project, the planned quantity of each type of machinery in the current construction project, and the planned operating time of each type of machinery in the current construction project.

[0035] Among them, the pre-planned construction projects refer to construction projects that have already been completed.

[0036] The term "current construction project" refers to construction projects that have not yet been completed.

[0037] Among them, the pre-planned construction site refers to a construction site that has already been completed.

[0038] The term "current construction site" refers to a construction site that has not yet been completed.

[0039] The bill of quantities for the current construction project refers to a detailed list of all the engineering contents and their corresponding quantities in the current construction project.

[0040] The construction organization design document for the current construction project refers to the comprehensive technical document concerning the overall construction deployment, construction methods, construction sequence, resource allocation, and site layout of the current construction project.

[0041] The current construction project schedule refers to the time arrangement and interrelationship of each construction procedure and stage in the current construction project, including the start time, completion time, duration of each procedure, and logical relationship between procedures.

[0042] The beneficial effects of the embodiments of this application are as follows: Firstly, a text feature extraction network is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the current construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the current construction project. An attention network is then used to fuse these features to obtain a fused feature vector. This fused feature vector is input into a trained large language model, which generates the predicted carbon emissions of the current construction project at the current construction site. This method can quickly generate carbon emission prediction results before construction begins, which shortens the time for obtaining predicted carbon emissions and improves the efficiency of obtaining predicted carbon emissions compared to traditional post-construction statistical methods, thus improving the timeliness of carbon emission assessment. Secondly, since the predicted carbon emissions of the current construction project at the current site are automatically completed by the trained large language model, there is no need to rely on manual on-site surveys and experience-based judgments. The entire prediction process is not affected by factors such as the subjective experience of operators, differences in judgment standards, and human negligence, thus ensuring the objectivity and stability of the predicted carbon emissions and improving the reliability of the predicted carbon emissions.

[0043] Please see Figure 3 , Figure 3 The implementation flowchart of S203 provided in the embodiments of this application is described in detail below: S301, input the planned material usage, mechanical parameters and planned construction period of the preset construction project into the attention network; S302 uses an attention network to fuse the planned material usage, mechanical parameters, and planned construction period of a pre-defined construction project to obtain a fused feature vector of the pre-defined construction project. The fused feature vector of the pre-defined construction project and the actual carbon emissions of the pre-defined construction project at the pre-defined construction site are combined to form a sample. S303: Different samples are combined into a dataset. The dataset is used to train the pre-set large language model. The loss value of the large language model on the dataset is obtained through the mean squared error loss function. The goal is to reduce the loss value. When the loss value of the large language model on the dataset no longer decreases, the training of the large language model is stopped and the trained large language model is saved.

[0044] In this embodiment, when the loss value of the large language model on the dataset no longer decreases, it means that the large language model has converged to the optimal state of the current dataset. At this time, training of the large language model is stopped and the trained large language model is saved. This can save computing resources, reduce training time, and improve training efficiency.

[0045] For the construction site carbon emission monitoring method described in the above embodiments, please refer to [link / reference]. Figure 4 , Figure 4 This is a schematic block diagram of a construction site carbon emission monitoring device provided in an embodiment of this application. Figure 4 The construction site carbon emission monitoring device 400 shown can be applied to, for example... Figure 1 The application scenario diagram shows electronic devices. The following section uses electronic devices as an example to illustrate this. Figure 4 The construction site carbon emission monitoring device 400 shown is described in detail. The construction site carbon emission monitoring device 400 may include a first acquisition module 401, an extraction module 402, a fusion module 403, a second acquisition module 404, and a generation module 405.

[0046] The first acquisition module 401 is used to acquire multi-source data of the preset construction project. The multi-source data of the preset construction project includes the bill of quantities of the preset construction project, the construction organization design document of the preset construction project, and the construction schedule of the preset construction project. The extraction module 402 is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the preset construction project using a text feature extraction network, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the preset construction project. The fusion module 403 is used to use an attention network to fuse the planned material usage, mechanical parameters, and planned construction period of a preset construction project to obtain a fusion feature vector of the preset construction project. The fusion feature vector of the preset construction project and the actual carbon emissions of the preset construction project at the preset construction site are combined into a sample. Different samples are combined into a dataset. The dataset is used to train the preset large language model and obtain the loss value of the large language model on the dataset. When the loss value of the large language model on the dataset no longer decreases, the training of the large language model is stopped and the trained large language model is saved. The second acquisition module 404 is used to send a data acquisition request for the current construction project to the database and receive multi-source data of the current construction project returned by the database according to the data acquisition request. The multi-source data of the current construction project includes the bill of quantities of the current construction project, the construction organization design document of the current construction project, and the construction schedule of the current construction project. The generation module 405 uses a text feature extraction network to extract features from the bill of quantities, construction organization design documents, and construction schedule of the current construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the current construction project. An attention network is then used to fuse these features to obtain a fused feature vector. This fused feature vector is input into a trained large language model, which then generates the predicted carbon emissions of the current construction project at the current construction site.

[0047] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0048] Please see Figure 5 , Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application.

[0049] like Figure 5 As shown, Figure 5 The electronic device includes: at least one processor 20, a memory 21, and a computer program 22 stored in the memory 21 and executable on the at least one processor 20, wherein the processor 20 executes the computer program 22 to implement the steps in any of the above method embodiments.

[0050] The electronic device may include, but is not limited to, processor 20 and memory 21. Those skilled in the art will understand that... Figure 5 This is merely an example of an electronic device and does not constitute a limitation on electronic devices. It may include more or fewer components than shown in the illustration, or combinations of certain components, or different components. For example, it may also include input / output devices, network access devices, etc.

[0051] The processor 20 is used to run a computer program 22 stored in the memory 21, and performs the following steps when executing the computer program 22: Acquire multi-source data for the preset construction project. The multi-source data for the preset construction project includes the bill of quantities, the construction organization design document, and the construction schedule of the preset construction project. A text feature extraction network is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the pre-defined construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the pre-defined construction project. An attention network is used to fuse the planned material usage, mechanical parameters, and planned construction period of a pre-defined construction project to obtain a fused feature vector. The fused feature vector and the actual carbon emissions of the pre-defined construction project at the pre-defined construction site are combined to form a sample. Different samples are combined to form a dataset. The pre-defined large language model is trained using the dataset, and the loss value of the large language model on the dataset is obtained. When the loss value of the large language model on the dataset no longer decreases, the training of the large language model is stopped, and the trained large language model is saved. Send a data acquisition request for the current construction project to the database, and receive multi-source data of the current construction project returned by the database based on the data acquisition request. The multi-source data of the current construction project includes the bill of quantities, the construction organization design documents, and the construction schedule of the current construction project. A text feature extraction network is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the current construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the current construction project. An attention network is then used to fuse these three data to obtain a fused feature vector. This fused feature vector is then input into a trained large language model, which is used to generate the predicted carbon emissions of the current construction project at the current construction site.

[0052] The processor 20 may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors, field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0053] In some embodiments, the memory 21 may be an internal storage unit of the electronic device, such as a hard disk or memory of the electronic device.

[0054] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for monitoring carbon emissions from construction sites based on multi-source data fusion, characterized in that, The construction site carbon emission monitoring method, applied to electronic devices, includes: Acquire multi-source data for the preset construction project. The multi-source data for the preset construction project includes the bill of quantities, the construction organization design document, and the construction schedule of the preset construction project. A text feature extraction network is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the pre-defined construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the pre-defined construction project. An attention network is used to fuse the planned material usage, mechanical parameters, and planned construction period of a pre-defined construction project to obtain a fused feature vector. The fused feature vector and the actual carbon emissions of the pre-defined construction project at the pre-defined construction site are combined to form a sample. Different samples are combined to form a dataset. The pre-defined large language model is trained using the dataset, and the loss value of the large language model on the dataset is obtained. When the loss value of the large language model on the dataset no longer decreases, the training of the large language model is stopped, and the trained large language model is saved. Send a data acquisition request for the current construction project to the database, and receive multi-source data of the current construction project returned by the database based on the data acquisition request. The multi-source data of the current construction project includes the bill of quantities, the construction organization design documents, and the construction schedule of the current construction project. A text feature extraction network is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the current construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the current construction project. An attention network is then used to fuse these three data to obtain a fused feature vector. This fused feature vector is then input into a trained large language model, which is used to generate the predicted carbon emissions of the current construction project at the current construction site.

2. The construction site carbon emission monitoring method according to claim 1, characterized in that, Obtain multi-source data for the pre-defined construction project. This multi-source data includes the bill of quantities, construction organization design documents, and construction schedule. The system sends connection requests to the file storage system via a communication network and receives response messages from the file storage system based on the connection requests. The communication network includes Ethernet and WIFI networks. When the response message is an authentication success message, the multi-source data of the preset construction project is retrieved from the file storage system. The multi-source data of the preset construction project includes the bill of quantities of the preset construction project, the construction organization design document of the preset construction project, and the construction schedule of the preset construction project.

3. The construction site carbon emission monitoring method according to claim 1, characterized in that, An attention network is used to fuse the planned material usage, mechanical parameters, and planned construction period of a pre-defined construction project, resulting in a fused feature vector. This fused feature vector, along with the actual carbon emissions of the project at the pre-defined construction site, forms a sample. Different samples are combined into a dataset, which is then used to train a pre-defined large language model. The loss value of the large language model on the dataset is obtained. Training is stopped when the loss value of the large language model on the dataset no longer decreases, and the trained large language model is saved, including: Input the planned material usage, mechanical parameters, and planned construction period of the pre-defined construction project into the attention network; An attention network is used to fuse the planned material usage, mechanical parameters, and planned construction period of a pre-designed construction project to obtain a fused feature vector of the pre-designed construction project. The fused feature vector of the pre-designed construction project and the actual carbon emissions of the pre-designed construction project at the pre-designed construction site are then combined to form a sample. Different samples are combined into a dataset, and the dataset is used to train a pre-defined large language model. The loss value of the large language model on the dataset is obtained through the mean squared error loss function. The goal is to reduce the loss value. When the loss value of the large language model on the dataset no longer decreases, the training of the large language model is stopped, and the trained large language model is saved.

4. The construction site carbon emission monitoring method according to claim 1, characterized in that, The text feature extraction network is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the current construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period. An attention network is then used to fuse these features to obtain a fused feature vector. This fused feature vector is input into a trained large language model. The trained large language model then generates the predicted carbon emissions of the current construction project at the current construction site. The site carbon emission monitoring method includes: When the predicted carbon emissions of the current construction project at the current site are not less than the preset value, emission reduction suggestions for the current construction project will be displayed. These suggestions include reducing the operating time of the target machinery and adjusting the construction sequence. If the predicted carbon emissions of the current construction project at the current site are less than the preset value, a message indicating that the current construction project has been approved will be displayed.

5. The construction site carbon emission monitoring method according to claim 1, characterized in that, After displaying a pass message for the current construction project when the predicted carbon emissions at the current construction site are less than a preset value, the construction site carbon emission monitoring method includes: Obtain monitoring parameters for the current construction project, including dust concentration, noise levels, and energy consumption. When the dust concentration exceeds the preset concentration, a start command is sent to the fog cannon; when the dust concentration drops back to the qualified range, a shut-off command is sent to the fog cannon. When the noise level exceeds the preset noise limit, noise reduction suggestions are output, including adjusting the process or replacing with low-noise equipment. When the noise level does not exceed the preset noise limit, a noise compliance message is output. When energy consumption exceeds the preset consumption, optimization suggestions are output, including shortening the single running time of the process equipment and delaying the start-up time of the process equipment to the off-peak electricity price period.

6. The construction site carbon emission monitoring method according to claim 5, characterized in that, Obtain monitoring parameters for the current construction project, including dust concentration, noise levels (decibels), and energy consumption. Obtain the monitoring time for the current construction project and determine whether the current time is a monitoring time. If the current time is the monitoring time, obtain the monitoring parameters of the current construction project, including dust concentration, noise decibels, and energy consumption.

7. The construction site carbon emission monitoring method according to claim 1, characterized in that, The pre-set mechanical parameters for the construction project include the rated power of each type of machinery in the pre-set construction project, the planned quantity of each type of machinery in the pre-set construction project, and the planned operating time of each type of machinery in the pre-set construction project.

8. The construction site carbon emission monitoring method according to claim 1, characterized in that, The mechanical parameters of the current construction project include the rated power of each type of machinery in the current construction project, the planned quantity of each type of machinery in the current construction project, and the planned operating time of each type of machinery in the current construction project.

9. A construction site carbon emission monitoring device based on multi-source data fusion, characterized in that, Applied to electronic devices, including: The first acquisition module is used to acquire multi-source data of the preset construction project. The multi-source data of the preset construction project includes the bill of quantities of the preset construction project, the construction organization design document of the preset construction project, and the construction schedule of the preset construction project. The extraction module is used to extract features from the bill of quantities, construction organization design documents, and construction schedule of the preset construction project using a text feature extraction network, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period of the preset construction project. The fusion module uses an attention network to fuse the planned material usage, mechanical parameters, and planned construction period of a pre-defined construction project to obtain a fused feature vector. This fused feature vector and the actual carbon emissions of the pre-defined construction project at the pre-defined construction site are combined to form a sample. Different samples are combined to form a dataset. The dataset is used to train a pre-defined large language model, and the loss value of the large language model on the dataset is obtained. When the loss value of the large language model on the dataset no longer decreases, the training of the large language model is stopped, and the trained large language model is saved. The second acquisition module is used to send a data acquisition request for the current construction project to the database and receive multi-source data of the current construction project returned by the database according to the data acquisition request. The multi-source data of the current construction project includes the bill of quantities of the current construction project, the construction organization design document of the current construction project, and the construction schedule of the current construction project. The generation module uses a text feature extraction network to extract features from the bill of quantities, construction organization design documents, and construction schedule of the current construction project, respectively, to obtain the planned material usage, mechanical parameters, and planned construction period. An attention network is then used to fuse these features to obtain a fused feature vector. This fused feature vector is input into a trained large language model, which then generates the predicted carbon emissions of the current construction project at the current construction site.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the construction site carbon emission monitoring method as described in any one of claims 1 to 7.

Images