A power grid infrastructure engineering project risk assessment method and system based on a GCN model

Abstract

The application discloses a kind of power grid infrastructure engineering project risk assessment method and system based on GCN model include, through power grid infrastructure project data, construction engineering project index system, and divide system structure, establish engineering project risk characteristic evaluation model;The data is cleaned, and the engineering project risk characteristic evaluation model is used as input, and a GCN model is constructed;The GCN model is trained and tested, and the risk level and model prediction result are output;The present application applies Lagrange interpolation method to the completion of power grid risk data, improves the accuracy and reliability of data;Model parameters are optimized using Adam optimizer and traditional gradient descent algorithm, reducing the error of output model prediction result;On the other hand, the risk influencing factors can be mined by the system and abstracted into a topological graph, effectively improving the risk prediction accuracy of power grid infrastructure construction projects, greatly reducing the error of manual judgment.

Description

Technical Field

[0001] This invention relates to the field of engineering project risk assessment technology, and in particular to a method and system for risk assessment of power grid infrastructure engineering projects based on the GCN model. Background Technology

[0002] Risk assessment for power grid infrastructure projects is a crucial step in the power industry, as it evaluates the risks associated with these projects and provides vital support for decision-making. However, in practice, risk assessment for power grid infrastructure projects faces numerous difficulties and challenges.

[0003] First, inadequate data management is a significant problem in the risk assessment of power grid infrastructure projects. These projects involve a vast amount of data, encompassing aspects such as project progress, quality, personnel management, and financial management. However, due to the dispersed data sources, inconsistent data formats, and varying data quality, data management becomes extremely difficult. Second, data mining is another challenge in the risk assessment of power grid infrastructure projects. The sheer volume of data from these projects is enormous, and the relationships between these data points are complex. Extracting effective information and uncovering potential patterns from this data is a crucial issue in the risk assessment of power grid infrastructure projects. Finally, the large margin of error in human judgment is also a challenge in the risk assessment of power grid infrastructure projects. Due to the massive volume of data and the complex relationships between them, traditional human judgment methods are prone to misjudgment, leading to inaccurate assessment results and posing risks to project decisions. Summary of the Invention

[0004] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0005] In view of the aforementioned existing problems, this invention is proposed. Therefore, this invention provides a risk assessment method for power grid infrastructure projects based on the GCN model, to address the problems of inadequate data management, difficulties in data mining, and large errors in manual judgment in practical applications.

[0006] To address the aforementioned technical problems, this invention provides the following technical solution: a risk assessment method for power grid infrastructure projects based on the GCN model, comprising:

[0007] By using data from power grid infrastructure projects, an indicator system for engineering projects is constructed, the system structure is divided, and a risk characteristic assessment model for engineering projects is established.

[0008] The data is cleaned, and the project risk characteristic assessment model is used as input to construct the GCN model;

[0009] The GCN model is trained and tested, and the risk level and model prediction results are output.

[0010] As a preferred embodiment of the risk assessment method for power grid infrastructure projects based on the GCN model described in this invention, the method involves: constructing an indicator system for the project using power grid infrastructure project data, and defining the system structure, including:

[0011] The project risk feature assessment model is established by dividing the project into four major categories: "human," "equipment," "management," and "environment," and using these categories as feature data. The model also uses the predicted risk level of the project as the input data for the GCN model.

[0012] Based on the type of project, engineering projects are divided into 7 categories: thermal power projects, hydropower projects, wind power projects, power line projects, photovoltaic projects, nuclear power projects, and power supply bureau facility projects.

[0013] As a preferred embodiment of the risk assessment method for power grid infrastructure projects based on the GCN model described in this invention, the data cleaning process includes:

[0014] Check the data for missing, abnormal, or duplicate values;

[0015] The data is completed using Lagrange interpolation. Assume the polynomial is f(k), and the values ​​at each point are (x...). i ,y i When calculating the coordinates of point k, the formula is as follows:

[0016] When the k-th data item is missing in the dataset, the solution is calculated for the n data items in the dataset.

[0017] Discretize the feature data in the risk feature assessment model for engineering projects.

[0018] As a preferred embodiment of the risk assessment method for power grid infrastructure projects based on the GCN model described in this invention, the discretization process includes:

[0019] If the construction workers possess the necessary safety knowledge for the operation, the value is 1; otherwise, it is 0. If the construction workers possess the required skills or knowledge, the value is 1; otherwise, it is 0. If the construction workers have the corresponding professional license, the value is 1; otherwise, it is 0. If the construction company has a certificate of qualified safety education and training for its employees, the value is 1; otherwise, it is 0. If the construction company has a qualified supervision and management model for employees' adherence to rules and regulations, the value is 1; otherwise, it is 0. If the construction company's management personnel are on duty as required during the operation, the value is 1; otherwise, it is 0. If the construction company's work instructions are accurate, the value is 1; otherwise, it is 0. If the safety equipment meets the actual work requirements, the value is 1; otherwise, it is 0. If the work equipment meets the actual work requirements, the value is 1; otherwise, it is 0. If the equipment's lifespan is greater than 50% of its design lifespan, the value is 1; otherwise, it is 0. If the site survey before the operation is accurate, the value is 1; otherwise, it is 0. If the management personnel supervise the operation on-site, the value is 1; otherwise, it is 0.

[0020] As a preferred embodiment of the risk assessment method for power grid infrastructure projects based on the GCN model described in this invention, the method involves: constructing a GCN model by taking the project risk characteristic assessment model as input, including:

[0021] Save the cleaned data again;

[0022] Using the torch method, the constructed engineering project indicator system is incorporated to transform the data into a graph structure; the cleaned data is then converted into graph structure data, and edges are generated for each data point with itself as a node.

[0023] As a preferred embodiment of the risk assessment method for power grid infrastructure projects based on the GCN model described in this invention, the cleaned data includes:

[0024] Safety knowledge, skill level or knowledge level, professional license, safety education and training certificate, supervision and management model, on-the-job performance under management, accuracy of work instructions, condition of safety equipment, condition of work equipment, equipment quality, accuracy of on-site survey before operation and supervision model under management.

[0025] As a preferred embodiment of the risk assessment method for power grid infrastructure projects based on the GCN model described in this invention, the method includes: training and testing the GCN model, comprising:

[0026] The parameters in all GCN models are randomly initialized, and the parameters are optimized using the gradient descent algorithm with the Adam optimizer.

[0027] The Adam optimizer is used to apply gradient descent algorithm to the parameters, as expressed by the following formula:

[0028]

[0029] Where α is the learning rate, g t Let β1 and β2 be the gradient of the current parameters, respectively, and let m be the hyperparameters of the first and second momentum. t For the past gradient, V t w is the momentum variable. t As weight;

[0030] A three-layer GCN model is constructed to perform feature aggregation and prediction on the graph structure data;

[0031] The feature aggregation and prediction include:

[0032] First, the nodes of the graph data structure in the dataset are encoded using the GCN model to obtain vector representations of the nodes. Second, the positive and negative samples in the dataset are used for supervised learning using the vector representations, and the inner product of the node pairs in the samples is obtained using the vectors. The loss value is calculated by combining the inner product with the predicted risk level of the project. Finally, the parameters are updated by backpropagation, and the risk level prediction result is output.

[0033] In supervised learning, negative samples need to be resampled in each round of model training.

[0034] A risk assessment system for power grid infrastructure projects based on the GCN model includes: a project indicator system module, a project classification module, and a project risk assessment model module.

[0035] The engineering project indicator system module constructs an engineering project indicator system using data from power grid infrastructure projects, divides the system structure, and establishes an engineering project risk characteristic assessment model.

[0036] The project classification module is used to construct the GCN model by taking the project risk characteristic assessment model as input.

[0037] The project risk characteristic assessment model module is used to train and test the GCN model and output the risk level and model prediction results.

[0038] As part of the GCN-based risk assessment system for power grid infrastructure projects described in this invention, it also includes:

[0039] The project indicator system module is specifically used to classify project indicators into four major categories: "human," "equipment," "management," and "environment," and to use these indicators as feature data. It also uses the predicted project risk level as the input data for the GCN model, establishing a project risk characteristic assessment model. Based on the type of project, projects are categorized as follows: thermal power projects, hydropower projects, wind power projects, power line projects, photovoltaic projects, nuclear power projects, and power supply bureau facility projects.

[0040] A computer device includes a memory and a processor, the memory storing a computer program, wherein the processor executes the computer program to implement any step of the method described above.

[0041] A computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any step of the above-described method.

[0042] Compared with existing technologies, the invention has the following advantages: Firstly, by applying Lagrange interpolation to complete power grid risk data, the accuracy and reliability of the data are improved. Secondly, the Adam optimizer is used to optimize model parameters and the traditional gradient descent algorithm, enabling more accurate weight calculation and reducing errors in the output model prediction results. This has a good application effect on risk assessment of power grid infrastructure projects. Thirdly, the system constructed using this invention can uncover the hidden spatial and temporal correlations between risk influencing factors and abstract the hidden patterns and characteristics in the risks of power grid engineering projects into a topological graph. This effectively improves the accuracy of risk prediction for power grid infrastructure construction projects, greatly reduces errors in manual judgment, and improves the efficiency of power grid infrastructure projects. Attached Figure Description

[0043] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0044] Figure 1 This is a flowchart illustrating the overall process of a risk assessment method for power grid infrastructure projects based on the GCN model, as described in one embodiment of the present invention.

[0045] Figure 2 This is a diagram illustrating the iterative process of the GCN model loss function in a risk assessment method for power grid infrastructure projects based on the GCN model, as described in an embodiment of the present invention.

[0046] Figure 3This diagram illustrates the iterative process of the GCN model accuracy function in a risk assessment method for power grid infrastructure projects based on the GCN model, as described in one embodiment of the present invention. Detailed Implementation

[0047] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0048] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0049] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0050] This invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of this invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not adhering to the usual scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of this invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.

[0051] Furthermore, in the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used solely for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In addition, the terms "first," "second," or "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0052] Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0053] Example 1

[0054] Reference Figure 1 This is the first embodiment of the present invention, which provides a risk assessment method for power grid infrastructure projects based on the GCN model, including:

[0055] S1. Construct an engineering project indicator system using data from power grid infrastructure projects, divide the system structure, and establish an engineering project risk characteristic assessment model.

[0056] Furthermore, the project is divided into four major categories of indicators: "human," "equipment," "management," and "environment," and these indicators are used as feature data. Label data that predicts the risk level of the project is used as the input data for the GCN model to establish a risk characteristic assessment model for the project.

[0057] Furthermore, based on the type of project, engineering projects are divided into 7 categories: thermal power projects, hydropower projects, wind power projects, power supply line projects, photovoltaic projects, nuclear power projects, and power supply bureau facility projects.

[0058] It should be noted that, based on the accident classification and the personal injury and property damage caused by the accident, the risk level of the engineering project is divided into five levels: low risk, low-medium risk, medium risk, medium-high risk, and high risk.

[0059] S2. Clean the project data and use the project risk characteristic assessment model as input to construct the GCN model;

[0060] Further, the data cleaning steps are as follows:

[0061] Check the data for missing, abnormal, or duplicate values;

[0062] The data is completed using Lagrange interpolation. Assume the polynomial is f(k), and the values ​​at each point are (x...). i ,y i When calculating the coordinates of point k, the formula is as follows:

[0063]

[0064] When the k-th data item is missing in the dataset, the solution is calculated for the n data items in the dataset.

[0065] Discretize the feature data in the risk feature assessment model for engineering projects;

[0066] It should be noted that, taking the power grid risk data used in this invention as an example, the risk values ​​are set as [5,4,1,1,missing,4,3,2], and the average value is 2.86; however, the value obtained by using the Lagrange interpolation method is 2.91, which improves the accuracy of the data; moreover, when faced with multiple missing data, such as [5,4,1,1,missing,4,3,2,3,2,4,5,missing,3], if the average value is used, the missing data in both places is 3.36, but the values ​​obtained by using the Lagrange interpolation method are 3 and 2 respectively; thus, it can be seen that the Lagrange interpolation method is more practical and reliable.

[0067] Further discretization processing includes:

[0068] If the construction workers possess the necessary safety knowledge for the operation, the value is 1; otherwise, it is 0. If the construction workers possess the required skills or knowledge, the value is 1; otherwise, it is 0. If the construction workers have the corresponding professional license, the value is 1; otherwise, it is 0. If the construction company has a certificate of qualified safety education and training for its employees, the value is 1; otherwise, it is 0. If the construction company has a qualified supervision and management model for employees' adherence to rules and regulations, the value is 1; otherwise, it is 0. If the construction company's management personnel are on duty as required during the operation, the value is 1; otherwise, it is 0. If the construction company's work instructions are accurate, the value is 1; otherwise, it is 0. If the safety equipment meets the actual work requirements, the value is 1; otherwise, it is 0. If the work equipment meets the actual work requirements, the value is 1; otherwise, it is 0. If the equipment's lifespan is greater than 50% of its design lifespan, the value is 1; otherwise, it is 0. If the site survey before the operation is accurate, the value is 1; otherwise, it is 0. If the management personnel supervise the operation on-site, the value is 1; otherwise, it is 0.

[0069] It should be noted that the cleaning data includes safety knowledge, skill level or knowledge level, professional license, safety education and training certificate, supervision and management model, on-site presence under management, accuracy of work instructions, condition of safety equipment, condition of work equipment, equipment quality, accuracy of on-site survey before operation and supervision model under management.

[0070] Furthermore, the GCN model is constructed, including:

[0071] Save the cleaned data again;

[0072] Using the torch method, the constructed engineering project indicator system is incorporated to transform the data into a graph structure; the cleaned data is then converted into graph structure data, generating a graph where each data point has itself as a node;

[0073] Furthermore, each data point is connected to twelve cleaned data items by graph edges according to the engineering project classification (thermal power, hydropower, wind power, transmission lines, photovoltaic projects, nuclear power, power supply bureaus), label (risk level 1-5), and graph nodes are added to each project. This transforms the purely numerical data into graph structure data, allowing data points that originally had no obvious correlation to generate graph edges by using themselves as nodes. The final graph data structure has 2047 nodes and 773542 edges.

[0074] S3. Train and test the GCN model, and output the risk level and model prediction results;

[0075] Furthermore, the training parameters are set, including setting the number of iterations to 1000 and the number of samples for each training iteration to 1.

[0076] All parameters in the GCN model are randomly initialized, and the parameters are optimized using the gradient descent algorithm with the Adam optimizer.

[0077] The Adam optimizer is used to apply gradient descent algorithm to the parameters, as expressed by the following formula:

[0078]

[0079] Where α is the learning rate, g t Let β1 and β2 be the gradient of the current parameters, respectively, and let m be the hyperparameters of the first and second momentum. t For the past gradient, V t w is the momentum variable. t As weight;

[0080] It should be noted that the learning rate α is initialized to 0.005, and the weights w t The decay rate was initialized to 0.0005; and the model was built using the DGL deployment message passing mechanism.

[0081] A three-layer GCN model is constructed to perform feature aggregation and prediction on the graph structure data;

[0082] The feature aggregation and prediction include:

[0083] First, the nodes of the graph data structure in the dataset are encoded using the GCN model to obtain vector representations of the nodes. Second, the positive and negative samples in the dataset are used for supervised learning using the vector representations, and the inner product of the node pairs in the samples is obtained using the vectors. The loss value is calculated by combining the inner product with the predicted risk level of the project. Finally, the parameters are updated by backpropagation, and the risk level prediction result is output.

[0084] In particular, supervised learning requires resampling negative samples in each round of GCN model training;

[0085] It should be noted that the iterative process is the same as the training process, which means that the model's loss function continuously decreases and its accuracy gradually converges; and CUDA semantics are used to reduce training time.

[0086] Furthermore, this embodiment also provides a risk assessment system for power grid infrastructure projects based on the GCN model, including:

[0087] The engineering project indicator system module is used to classify engineering projects into four major categories: "human," "equipment," "management," and "environment," and uses the indicator system as feature data; the module also uses the label data for predicting the risk level of engineering projects as input data for the GCN model.

[0088] The project classification module is used to categorize engineering projects into thermal power projects, hydropower projects, wind power projects, power line projects, photovoltaic projects, nuclear power projects, and power supply bureau facility projects.

[0089] The engineering project risk assessment model module is used to construct engineering project entity nodes from engineering project features and labels; add relationships between nodes to engineering project entities using engineering project classification to construct a graph structure with engineering project entities as nodes; and use the graph structure as input data, graph node data as input features, and risk level as prediction label to train the GCN model.

[0090] Furthermore, the project indicator system module also includes a data cleaning unit;

[0091] The data cleaning unit is used to clean the input feature data before training the GCN model and to check for missing, abnormal, or duplicate values ​​in the data.

[0092] Furthermore, the project classification module also includes a graph data classification unit;

[0093] The data classification unit is used to refine the project categories after the engineering project has been divided.

[0094] Furthermore, the project risk characteristic assessment model module also includes a graph data modeling unit;

[0095] The graph data modeling unit uses graph node data as input features, risk level as prediction level, and project classification as relationship to construct graph structure data.

[0096] This embodiment also provides a computer device applicable to risk assessment methods for power grid infrastructure projects based on the GCN model, including:

[0097] The system includes a memory and a processor. The memory stores computer-executable instructions, and the processor executes these instructions to implement the risk assessment method for power grid infrastructure projects based on the GCN model proposed in the above embodiments.

[0098] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0099] This embodiment also provides a storage medium storing a computer program that, when executed by a processor, implements the risk assessment method for power grid infrastructure projects based on the GCN model proposed in the above embodiments.

[0100] The storage medium proposed in this embodiment and the data storage method proposed in the above embodiments belong to the same inventive concept. Technical details not described in detail in this embodiment can be found in the above embodiments, and this embodiment has the same beneficial effects as the above embodiments.

[0101] Example 2

[0102] Reference Figures 2 to 3 This is the second embodiment of the present invention, which provides a risk assessment method for power grid infrastructure projects based on the GCN model, including:

[0103] The beneficial effects of the method of the present invention are explained and compared through simulation experiments and tabular data.

[0104] Taking into account both time and effectiveness, this invention selects 1000 iterations.

[0105] Depend on Figure 2 As shown, during the training process of the GCN model on the data, the loss value decreases with the increase of the number of iterations. When the number of iterations reaches 200, the loss value drops significantly and then gradually flattens out.

[0106] Depend on Figure 3 As shown, during the training process of the GCN model on the data, the accuracy increases with the number of iterations. When the number of iterations reaches 200, the accuracy improvement curve begins to flatten out.

[0107] Table 1

[0108]

[0109]

[0110] As shown in Table 1, the loss value and accuracy of the GCN model change continuously with the increase of the number of iterations during the training process. Table 1 shows the changes in loss value and accuracy within 1000 iterations. Since the increase in the number of iterations leads to an increase in model training time, the model training method of this invention achieves a loss value of 0.5313 and an accuracy of 0.8569.

[0111] Therefore, the method of the present invention can bring unexpected benefits. By using the Adam optimizer to optimize the model parameters and the traditional gradient descent algorithm, the weights can be calculated more accurately, reducing the error of the GCN model output prediction results. It has a good application effect on the risk assessment of power grid infrastructure projects.

[0112] Those skilled in the art will understand that embodiments of the present invention 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 implemented 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. The solutions in the embodiments of this application can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0113] 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 A device that provides the functions specified in one or more boxes.

[0114] These 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 specified in one or more boxes.

[0115] 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 specified in one or more boxes.

[0116] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0117] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A risk assessment method for power grid infrastructure projects based on the GCN model, characterized in that, include: By using data from power grid infrastructure projects, an indicator system for engineering projects is constructed, the system structure is divided, and a risk characteristic assessment model for engineering projects is established. The GCN model is constructed using the project risk characteristic assessment model as input. The GCN model is trained and tested, and the risk level and model prediction results are output. When it is determined that the data contains missing or duplicate values, Lagrange interpolation is used to complete the data. Suppose the polynomial is f(k), and the values ​​at each point are... When calculating the coordinates of point k, the formula is as follows: When the k-th data item is missing in the dataset, the solution is calculated for the n data items in the dataset. Discretize the feature data in the risk feature assessment model for engineering projects; Using the aforementioned project risk characteristic assessment model as input, a GCN model is constructed, including: Save the cleaned data again; The cleaned data includes: Safety knowledge, skill level or knowledge level, professional license, safety education and training certificate, supervision and management model, on-the-job performance under management, accuracy of work instructions, condition of safety equipment, condition of work equipment, equipment quality, accuracy of on-site survey before work and supervision model under management. Using the torch method, the constructed engineering project indicator system is incorporated to transform the data into a graph structure; the cleaned data is then converted into graph structure data, and edges of the graph with each data point as its node are generated. Training and testing the GCN model includes: Set the training parameters, including setting the number of iterations to 1000 and the number of samples per training session to 1. All parameters in the GCN model are randomly initialized, and the parameters are optimized using the gradient descent algorithm with the Adam optimizer. The Adam optimizer is used to apply gradient descent algorithm to the parameters, as expressed by the following formula: in, For learning rate, The gradient of the current parameters. and These are the hyperparameters for first-order momentum and second-order momentum, respectively. For the past gradient, For momentum variables, As weight; A three-layer GCN model is constructed to perform feature aggregation and prediction on the graph structure data; The feature aggregation and prediction include: First, the nodes of the graph data structure in the dataset are encoded using the GCN model to obtain vector representations of the nodes. Second, the positive and negative samples in the dataset are used for supervised learning using the vector representations, and the inner product of the node pairs in the samples is obtained using the vectors. The loss value is calculated by combining the inner product with the predicted risk level of the project. Finally, the parameters are updated by backpropagation, and the risk level prediction result is output. In particular, supervised learning requires resampling negative samples in each round of GCN model training.

2. The risk assessment method for power grid infrastructure projects based on the GCN model as described in claim 1, characterized in that, Using data from power grid infrastructure projects, an engineering project indicator system is constructed, and its structure is divided, including: The project risk feature assessment model is established by dividing the project into four major categories: "human factors", "equipment", "management" and "environment" and using the indicator system as feature data; the project risk level is predicted as the label data input to the GCN model. Based on the type of project, engineering projects are divided into the following categories: thermal power projects, hydropower projects, wind power projects, power line projects, photovoltaic projects, nuclear power projects, and power supply bureau facility projects.

3. The risk assessment method for power grid infrastructure projects based on the GCN model as described in claim 2, characterized in that, The discretization process includes: If the construction workers possess the necessary safety knowledge for the operation, the value is 1; otherwise, it is 0. If the construction workers possess the required skills or knowledge, the value is 1; otherwise, it is 0. If the construction workers have the corresponding professional license, the value is 1; otherwise, it is 0. If the construction company has a certificate of qualified safety education and training for its employees, the value is 1; otherwise, it is 0. If the construction company has a qualified supervision and management model for employees' adherence to rules and regulations, the value is 1; otherwise, it is 0. If the construction company's management personnel are on duty as required during the operation, the value is 1; otherwise, it is 0. If the construction company's work instructions are accurate, the value is 1; otherwise, it is 0. If the safety equipment meets the actual work requirements, the value is 1; otherwise, it is 0. If the work equipment meets the actual work requirements, the value is 1; otherwise, it is 0. If the equipment's lifespan is greater than 50% of its design lifespan, the value is 1; otherwise, it is 0. If the site survey before the operation is accurate, the value is 1; otherwise, it is 0. If the management personnel supervise the operation on-site, the value is 1; otherwise, it is 0.

4. A risk assessment system for power grid infrastructure projects based on the GCN model, characterized in that, The method of claim 1 is used, comprising: Engineering project indicator system module, engineering project classification module, engineering project risk characteristic assessment model module; The engineering project indicator system module constructs an engineering project indicator system using data from power grid infrastructure projects, divides the system structure, and establishes an engineering project risk characteristic assessment model. The project classification module is used to construct the GCN model by taking the project risk characteristic assessment model as input. The project risk characteristic assessment model module is used to train and test the GCN model and output the risk level and model prediction results.

5. The risk assessment system for power grid infrastructure projects based on the GCN model as described in claim 4, characterized in that, The project indicator system module is specifically used to classify project indicators into four major categories: "human," "equipment," "management," and "environment," and to use these indicators as feature data. It also uses labeled data, which is the input data for the GCN model, to establish a project risk characteristic assessment model. Based on the type of project, projects are categorized as follows: thermal power projects, hydropower projects, wind power projects, power line projects, photovoltaic projects, nuclear power projects, and power supply bureau facility projects.

6. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the method according to any one of claims 1 to 3.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1 to 3.

Images