Electric power project risk prediction method and system based on data mining

Abstract

The invention discloses an electric power project risk prediction method and system based on data mining, and relates to the technical field of electric power project risk prediction.The method comprises the steps that electric power project data are collected, the electric power project data are preprocessed, key features influencing project risks are extracted, and feature vectors are established; constructing an electric power project risk prediction model, and performing risk prediction analysis on the electric power project based on the feature vector; and taking a part of areas with risks as nodes to form a topological graph, identifying key paths and bottlenecks, and formulating corresponding resource allocation strategies. According to the method, the electric power project risk prediction model is constructed based on the feature vectors, risk prediction analysis is carried out on the electric power project in combination with the time change trend, the accuracy and foresight of risk prediction are improved, the topological graph is formed by taking part of areas with risks as nodes, key paths and bottlenecks are identified, and the risk prediction accuracy is improved. And a corresponding resource allocation strategy is formulated, so that the security and the stability of the project are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power project risk prediction, and in particular to a power project risk prediction method and system based on data mining. Background Art

[0002] Risk management and prediction of power projects have always been an important research direction in the power industry. With the continuous growth of electricity demand and the expansion of power grid scale, the complexity and uncertainty of power projects have also increased. Traditional risk management methods mainly rely on experience and expert judgment. Although they can effectively identify and mitigate risks to a certain extent, their limitations are gradually revealed in the face of complex and dynamically changing power project environments. In recent years, with the rapid development of big data and artificial intelligence technologies, the application of data mining technology in power projects has gradually received attention. Data mining technology can effectively extract the key factors affecting project risks by analyzing and processing a large amount of historical data, and establish prediction models based on these factors, providing scientific basis and technical support for risk management of power projects.

[0003] Although data mining technology has broad application prospects in power project risk management, the existing technology still has some shortcomings. The existing data preprocessing methods are often unable to fully clean and standardize multi-source heterogeneous data, resulting in data quality problems in the feature extraction process, which in turn affects the accuracy of the prediction model. Most of the existing feature extraction methods are based on simple statistical analysis and lack in-depth mining of complex nonlinear relationships, resulting in the extracted features cannot fully reflect the key factors affecting project risks. In addition, most of the current risk prediction models are based on linear regression or traditional machine learning algorithms, and fail to fully utilize the advantages of algorithms such as deep learning, resulting in insufficient accuracy and robustness of the prediction results. The existing risk management methods lack analysis of the spatial distribution characteristics of power project risks, cannot effectively identify key paths and bottlenecks, and are difficult to formulate targeted resource allocation strategies. Summary of the invention

[0004] In view of the above-mentioned problems, the present invention is proposed.

[0005] Therefore, the technical problem solved by the present invention is: the existing power project risk prediction methods are unable to fully clean and standardize multi-source heterogeneous data, cannot deeply explore complex nonlinear relationships, the accuracy and robustness of the prediction model are insufficient, and there are problems such as how to effectively identify the key paths and bottlenecks of power project risks.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a method for predicting power project risks based on data mining, comprising collecting power project data, preprocessing the power project data, extracting key features that affect project risks, and establishing feature vectors; constructing a power project risk prediction model, and performing risk prediction analysis on the power project based on the feature vectors; forming a topological map with some risky areas as nodes, identifying critical paths and bottlenecks, and formulating corresponding resource allocation strategies.

[0007] As a preferred solution of the power project risk prediction method based on data mining described in the present invention, the preprocessing includes collecting power project data, including power consumption data, weather data, equipment operation data, and geographic information data; performing data cleaning and data standardization on the power project data, wherein the data cleaning includes deleting outliers, and the process of data standardization is expressed as follows:

[0008]

[0009] Among them, X' i is the standardized power project data, X i is the original data of the ith power project, X i ∈{P i ,W i ,E i ,G i}, P i is the power consumption data, W i is weather data, E i G is the equipment operation data, i is geographic information data, σ X is the standard deviation of the power project data set X, μ X is the mean of the power project data set X.

[0010] As a preferred solution of the power project risk prediction method based on data mining described in the present invention, wherein: the establishment of the feature vector includes feature selection based on LASSO regression, extracting key features from the power project data, and calculating the weight of each feature It is expressed as:

[0011]

[0012] in, represents the weight of each feature, β is the regression coefficient vector, which represents the variable in the feature selection process, n is the number of samples, which represents the number of power projects, and y i represents the risk value of the ith power project, p represents the number of features extracted from the power project data, and β jis the jth coefficient in the regression coefficient vector, indicating the weight of feature j, λ is the regularization parameter, X i represents the feature vector of the ith power project, is the nonlinear transformation of the jth feature of the i-th power project, expressed as:

[0013]

[0014] Where m represents the number of all original features in the power project dataset, α jk represents the coefficient between feature j and the kth original data feature, X ik Represents the kth original data feature of the ith power project.

[0015] As a preferred solution of the power project risk prediction method based on data mining described in the present invention, wherein: the construction of the power project risk prediction model includes combining the time change trend, based on the characteristics of the power project risk changing over time, to construct the objective function of the power project risk prediction model, which is expressed as:

[0016]

[0017] Among them, w(t) is the weight vector in the time dimension, b(t) is the bias in the time dimension, ξ(t) is the slack variable in the time dimension, C is the penalty intensity for controlling the slack variable, ξ i (t) represents the slack variable of the i-th power project at time t; the constraint condition is expressed as:

[0018]

[0019] in, represents the predicted value of the power project risk, γ represents the kernel function parameter, X l is the characteristic vector of the lth power project; the weight vector on the time dimension is updated by the gradient descent method, and the optimal solution is gradually introduced in combination with the time change trend, and the weight vector on the time dimension is updated, which is expressed as:

[0020]

[0021] Among them, w(t-1) is the weight vector at time t-1, η is the learning rate, L represents the loss function, and the loss function is expressed as:

[0022]

[0023] The loss function is used to balance model complexity and prediction error.

[0024] As a preferred solution of the power project risk prediction method based on data mining described in the present invention, the risk prediction analysis includes setting the risk level based on the predicted value of the power project risk; when the predicted value of the power project risk is When the predicted value of the power project risk is When the predicted value of the power project risk is When the power project is involved in the area of ​​high risk.

[0025] As a preferred solution of the power project risk prediction method based on data mining described in the present invention, the forming of the topological graph includes taking high-risk and medium-risk areas as nodes, forming a weighted graph of power project areas and connections, and the weight of each edge represents the predicted risk value. The risk data is displayed in a graphical manner, which is represented as follows:

[0026] G=(V,ε,W)

[0027]

[0028] Among them, G is the formed weighted graph, representing different areas and connections in the power project, V is the node set, representing different areas in the power project, ε is the edge set, representing the connection between areas in the power project, and W(u,v) is the weight set of the edge from the intermediate node u to the target node v, which is determined by the predicted value of the power project risk; through data mining, combined with the connection risk of the area in the power project, the critical path and bottleneck area in the project are identified, and the connection risk analysis is expressed as:

[0029] d(v)=min{d(u)+W(u,v)|u∈V}

[0030] Among them, d(v) represents the cumulative risk value from the starting point to the target node v in the power project, and d(u) represents the cumulative risk value from the starting point to the intermediate node u in the power project.

[0031] As a preferred solution of the power project risk prediction method based on data mining described in the present invention, wherein: the formulation of the corresponding resource allocation strategy includes setting the risk threshold d max , identify the critical paths and bottlenecks in the power project and formulate corresponding resource allocation strategies; when d(v)≥d max When the corresponding edges and nodes are critical paths and bottlenecks, professional maintenance and repair personnel are dispatched to conduct timely inspection and maintenance of equipment and facilities in high-risk areas, upgrade or replace old equipment, add redundant equipment and backup systems at key nodes and paths, and switch to backup equipment when a failure occurs to maintain power supply.

[0032] Another object of the present invention is to provide a power project risk prediction system based on data mining, which can solve the problem of low reliability of current power project risk prediction by constructing a power project risk prediction model and conducting risk prediction analysis on power projects based on feature vectors.

[0033] As a preferred solution of the power project risk prediction system based on data mining described in the present invention, it includes: a feature extraction module, a risk prediction module, and a resource allocation module; the feature extraction module is used to collect power project data, perform power project data preprocessing, extract key features that affect project risks, and establish feature vectors; the risk prediction module is used to construct a power project risk prediction model, and perform risk prediction analysis on power projects based on feature vectors; the resource allocation module is used to form a topological map with some risky areas as nodes, identify critical paths and bottlenecks, and formulate corresponding resource allocation strategies.

[0034] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement a step of a method for predicting power project risks based on data mining.

[0035] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of a method for predicting risks of power projects based on data mining.

[0036] Beneficial effects of the present invention: The power project risk prediction method based on data mining provided by the present invention collects power project data, including power consumption data, weather data, equipment operation data and geographic information data, and performs data cleaning and standardization processing to establish feature vectors, thereby achieving comprehensive cleaning and standardization of multi-source heterogeneous data, eliminating data noise and scale differences, thereby improving data quality and consistency, and improving the accuracy and robustness of data analysis. By constructing a power project risk prediction model based on feature vectors and combining time-varying trends, risk prediction analysis of power projects is performed, thereby improving the accuracy and foresight of risk prediction, and achieving efficient prediction and dynamic monitoring of power project risks. By forming a topological map by taking some risky areas as nodes, identifying critical paths and bottlenecks, and formulating corresponding resource allocation strategies, high-risk paths and nodes can be effectively identified, and measures such as additional professional maintenance personnel and equipment upgrades can be taken in a timely manner to intervene, thereby reducing the overall risk of power projects, achieving precise risk management and optimal resource allocation for power projects, and improving the safety and stability of projects. The present invention achieves better results in terms of accuracy, reliability and efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0038] Figure 1 An overall flow chart of a method for predicting power project risks based on data mining is provided for the first embodiment of the present invention.

[0039] Figure 2 An overall flow chart of a power project risk prediction system based on data mining provided in the third embodiment of the present invention. DETAILED DESCRIPTION

[0040] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0041] Example 1, reference Figure 1 , as an embodiment of the present invention, provides a power project risk prediction method based on data mining, comprising:

[0042] S1: Collect power project data, pre-process the power project data, extract key features that affect project risks, and establish feature vectors.

[0043] Furthermore, preprocessing includes collecting power project data, including power consumption data, weather data, equipment operation data, and geographic information data; cleaning and standardizing the power project data. Data cleaning includes deleting outliers. The process of data standardization is expressed as follows:

[0044]

[0045] Among them, X′ i is the standardized power project data, X i is the original data of the ith power project, X i ∈{P i , W i , E i , G i}, P i is the power consumption data, W i is weather data, E i G is the equipment operation data,i is geographic information data, σ X is the standard deviation of the power project data set X, μ X is the mean of the power project data set X.

[0046] It should be noted that the establishment of feature vectors includes feature selection based on LASSO regression, extracting key features from power project data, and calculating the weight of each feature. It is expressed as:

[0047]

[0048] in, represents the weight of each feature, β is the regression coefficient vector, which represents the variable in the feature selection process, n is the number of samples, which represents the number of power projects, and y i represents the risk value of the ith power project, p represents the number of features extracted from the power project data, and β j is the jth coefficient in the regression coefficient vector, indicating the weight of feature j, λ is the regularization parameter, X i represents the feature vector of the ith power project, is the nonlinear transformation of the jth feature of the i-th power project, expressed as:

[0049]

[0050] Where m represents the number of all original features in the power project dataset, α jk represents the coefficient between feature j and the kth original data feature, X ik Represents the kth original data feature of the ith power project.

[0051] It should also be noted that standardization ensures the comparability between different types of data, and feature selection further screens out features that have a significant impact on risk, providing high-quality input data for subsequent risk prediction models, solving the problems of data noise and feature redundancy in traditional methods, and making model training more efficient and accurate.

[0052] S2: Construct a risk prediction model for power projects and conduct risk prediction analysis on power projects based on feature vectors.

[0053] Furthermore, constructing a power project risk prediction model includes combining the time change trend and constructing the objective function of the power project risk prediction model based on the characteristics of power project risk changing over time, which is expressed as:

[0054]

[0055] Among them, w(t) is the weight vector in the time dimension, b(t) is the bias in the time dimension, ξ(t) is the slack variable in the time dimension, C is the penalty intensity for controlling the slack variable, ξ i (t) represents the slack variable of the i-th power project at time t; the constraint condition is expressed as:

[0056]

[0057] in, represents the predicted value of the power project risk, γ represents the kernel function parameter, and X1 is the feature vector of the first power project. The weight vector on the time dimension is updated by the gradient descent method, and the optimal solution is gradually introduced in combination with the time change trend. The weight vector on the time dimension is updated, which is expressed as:

[0058]

[0059] Among them, w(t-1) is the weight vector at time t-1, η is the learning rate, L represents the loss function, and the loss function is expressed as:

[0060]

[0061] The loss function is used to balance model complexity and prediction error.

[0062] It should be noted that risk forecasting analysis includes setting risk levels based on the predicted value of power project risks; When the predicted value of the power project risk is When the predicted value of the power project risk is When the power project is involved in the area of ​​high risk.

[0063] It should also be noted that by dynamically adjusting the model parameters, the model can accurately predict risks in different time periods, which makes up for the defect that traditional static models cannot capture changes over time, making risk prediction more comprehensive and reliable.

[0064] S3: Use some risky areas as nodes to form a topological map, identify critical paths and bottlenecks, and formulate corresponding resource allocation strategies.

[0065] Furthermore, forming a topological graph includes taking high-risk and medium-risk areas as nodes, forming a weighted graph of power project areas and connections, and the weight of each edge represents the predicted risk value. The risk data is displayed in a graphical way, which is expressed as:

[0066] G=(V,ε,W)

[0067]

[0068] Among them, G is the formed weighted graph, representing different areas and connections in the power project, V is the node set, representing different areas in the power project, ε is the edge set, representing the connection between areas in the power project, and W(u, v) is the weight set of the edge from the intermediate node u to the target node v, which is determined by the predicted value of the power project risk; through data mining, combined with the connection risk of the area in the power project, the critical path and bottleneck area in the project are identified, and the analysis of the connection risk is expressed as:

[0069] d(v)=min{d(u)+W(u,v)|u∈V}

[0070] Among them, d(v) represents the cumulative risk value from the starting point to the target node v in the power project, and d(u) represents the cumulative risk value from the starting point to the intermediate node u in the power project.

[0071] It should be noted that formulating the corresponding resource allocation strategy includes setting the risk threshold d max , identify the critical paths and bottlenecks in the power project and formulate corresponding resource allocation strategies; when d(v)≥d max When the corresponding edges and nodes are critical paths and bottlenecks, professional maintenance and repair personnel are dispatched to conduct timely inspection and maintenance of equipment and facilities in high-risk areas, upgrade or replace old equipment, add redundant equipment and backup systems at key nodes and paths, and switch to backup equipment when a failure occurs to maintain power supply.

[0072] It should also be noted that the topology map intuitively displays the connection risks of various areas in the power project, allowing managers to clearly identify high-risk areas and paths, and take timely measures to intervene. Compared with traditional methods, the present invention is more targeted and predictive in risk management, and can effectively reduce the overall risk of the project and improve resource utilization efficiency.

[0073] Example 2, an embodiment of the present invention, provides a method for predicting risk of power projects based on data mining. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0074] Firstly, the power consumption data, weather data, equipment operation data and geographic information data of a power project are selected as sample data. The sample data covers 6 months and involves 6 major power areas. The data of each area include power consumption, average temperature, number of equipment failures and regional population density. The data are obtained from the power company database, meteorological department and equipment maintenance records. The data of 6 power areas are collected, including power consumption data (kWh), weather data (℃), equipment operation data (number of failures) and geographic information data (population density). Outliers are deleted, such as obvious false positives in the number of equipment failures and erroneous values ​​in data collection. LASSO regression is used to extract key features from the standardized data. Power consumption, temperature, number of equipment failures and population density are selected as features. The weight of each feature is calculated. Based on the extracted features and combined with the time change trend, a power project risk prediction model is constructed. The risk level is set based on the predicted value of the power project risk. The high-risk and medium-risk areas are used as nodes to form a weighted graph. Through data mining, combined with the connection risk of the region, the critical path and bottleneck are identified, and the corresponding resource allocation strategy is formulated. Refer to Table 1 to record and analyze the experimental data.

[0075] Table 1 Experimental data record table

[0076]

[0077] Through the data, we can analyze the advantages of the method of the present invention in the risk prediction and management of power projects. The data preprocessing step improves the data quality. Through data cleaning and standardization, the noise and scale differences in the original data are eliminated, ensuring the consistency and reliability of the data. The feature extraction and feature vector establishment steps identify the key factors affecting the risk of power projects. The LASSO regression method is used to extract key features such as power consumption, temperature, number of equipment failures and population density from the standardized data, and their weights are calculated. Through the calculation of feature weights, the complex nonlinear relationship between each feature and project risk can be deeply explored, the accuracy of feature selection is improved, and it is ensured that the input of the risk prediction model has high correlation and high information content. The step of constructing a risk prediction model improves the accuracy and foresight of risk prediction by combining the time change trend. The model constructed based on the feature vector can dynamically capture the risk of power. The model optimizes the weight vector through the gradient descent method and gradually approaches the optimal solution, thereby realizing accurate prediction of project risks. By setting the risk level, the risk levels of different regions can be clearly distinguished, providing a scientific basis for risk management, forming a topological map and identifying the steps of critical paths and bottlenecks. By taking high-risk and medium-risk areas as nodes, a weighted graph is formed, and the critical paths and bottleneck areas are identified using the shortest path algorithm. The connection risks of various regions in the power project can be comprehensively analyzed, and the spatial distribution characteristics of the risks can be intuitively displayed. By setting the risk threshold, high-risk paths and nodes can be effectively identified, and measures such as additional maintenance personnel and equipment upgrades can be taken in time to intervene, thereby reducing the overall risk of the power project. The method of the present invention improves the scientificity and accuracy of power project risk management through comprehensive data preprocessing, feature extraction methods and accurate risk prediction models.

[0078] Example 3, reference Figure 2 , as an embodiment of the present invention, provides a power project risk prediction system based on data mining, including a feature extraction module, a risk prediction module, and a resource allocation module.

[0079] The power project data is collected, preprocessed, key features that affect project risks are extracted, and feature vectors are established; the risk prediction module is used to build a risk prediction model for power projects and conduct risk prediction analysis on power projects based on feature vectors; the resource allocation module is used to form a topological map with some risky areas as nodes, identify key paths and bottlenecks, and formulate corresponding resource allocation strategies

[0080] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0081] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0082] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0083] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc. It should be noted that the above embodiments are only used to illustrate the technical solution of the present invention and are not limited. Although the present invention is described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention, which should be included in the scope of the claims of the present invention.

[0084] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for predicting power project risks based on data mining, It is characterized in that include: Collect power project data, pre-process the power project data, extract key features that affect project risks, and establish feature vectors; Construct a risk prediction model for power projects and conduct risk prediction analysis on power projects based on feature vectors; Some risky areas are used as nodes to form a topological map, identify critical paths and bottlenecks, and formulate corresponding resource allocation strategies.

2. The power project risk prediction method based on data mining as claimed in claim 1, Features: The preprocessing includes collecting power project data, including power consumption data, weather data, equipment operation data, and geographic information data; Data cleaning and data standardization are performed on power project data. Data cleaning includes deleting outliers. The process of data standardization is expressed as follows: Among them, X' i is the standardized power project data, X i is the original data of the ith power project, X i ∈{P i ,W i ,E i ,G i }, P i is the power consumption data, W i is weather data, E i G is the equipment operation data, i is geographic information data, σ X is the standard deviation of the power project data set X, μ X is the mean of the power project data set X.

3. The power project risk prediction method based on data mining as claimed in claim 2, Features: The establishment of the feature vector includes feature selection based on LASSO regression, extracting key features from the power project data, and calculating the weight of each feature. It is expressed as: in, represents the weight of each feature, β is the regression coefficient vector, which represents the variable in the feature selection process, n is the number of samples, which represents the number of power projects, and y i represents the risk value of the ith power project, p represents the number of features extracted from the power project data, and β j is the jth coefficient in the regression coefficient vector, indicating the weight of feature j, λ is the regularization parameter, X i represents the feature vector of the ith power project, is the nonlinear transformation of the jth feature of the i-th power project, expressed as: Where m represents the number of all original features in the power project dataset, α jk represents the coefficient between feature j and the kth original data feature, X ik Represents the kth original data feature of the ith power project.

4. The power project risk prediction method based on data mining as claimed in claim 3, Features: The construction of the power project risk prediction model includes combining the time change trend and based on the characteristics of the power project risk changing over time, constructing the objective function of the power project risk prediction model, which is expressed as: Among them, w(t) is the weight vector in the time dimension, b(t) is the bias in the time dimension, ξ(t) is the slack variable in the time dimension, C is the penalty intensity for controlling the slack variable, ξ i (t) represents the slack variable of the ith power project at time t; The constraints are expressed as: in, represents the predicted value of power project risk, γ represents the kernel function parameter, X l is the characteristic vector of the lth power project; The weight vector on the time dimension is updated by the gradient descent method. Combined with the time change trend, the optimal solution is gradually introduced to update the weight vector on the time dimension, which is expressed as: Among them, w(t-1) is the weight vector at time t-1, η is the learning rate, L represents the loss function, and the loss function is expressed as: The loss function is used to balance model complexity and prediction error.

5. The power project risk prediction method based on data mining as claimed in claim 4, Features: The risk prediction analysis includes setting a risk level based on the predicted value of the power project risk; When the predicted value of power project risk When the power project involves an area with low risk; When the predicted value of power project risk When the power project involves an area with medium risk; When the predicted value of power project risk When the power project is involved in the area of ​​high risk.

6. The power project risk prediction method based on data mining as claimed in claim 5, Features: The forming of the topological graph includes taking high-risk and medium-risk areas as nodes, forming a weighted graph of power project areas and connections, the weight of each edge represents the predicted risk value, and displaying the risk data in a graphical manner, which is represented as follows: G=(V,ε,W) Among them, G is the formed weighted graph, representing different areas and connections in the power project, V is the node set, representing different areas in the power project, ε is the edge set, representing the connection between areas in the power project, and W(u,v) is the weight set of the edge from the intermediate node u to the target node v, which is determined by the predicted value of the power project risk; Through data mining, combined with the regional connection risks in the power project, the critical paths and bottleneck areas in the project are identified, and the connection risk analysis is expressed as: d(v)=min{d(u)+W(u,v)∣u∈V} Among them, d(v) represents the cumulative risk value from the starting point to the target node v in the power project, and d(u) represents the cumulative risk value from the starting point to the intermediate node u in the power project.

7. The power project risk prediction method based on data mining as claimed in claim 6, Features: The formulation of the corresponding resource allocation strategy includes setting a risk threshold d max , identify critical paths and bottlenecks in power projects and formulate corresponding resource allocation strategies; When d(v)≥d max When the corresponding edges and nodes are critical paths and bottlenecks, professional maintenance and repair personnel are dispatched to conduct timely inspection and maintenance of equipment and facilities in high-risk areas, upgrade or replace old equipment, add redundant equipment and backup systems at key nodes and paths, and switch to backup equipment when a failure occurs to maintain power supply.

8. A system using the power project risk prediction method based on data mining as claimed in any one of claims 1 to 7, Features: Including feature extraction module, risk prediction module, and resource allocation module; The feature extraction module is used to collect power project data, perform power project data preprocessing, extract key features that affect project risks, and establish feature vectors; The risk prediction module is used to construct a risk prediction model for power projects and to perform risk prediction analysis on power projects based on feature vectors; The resource allocation module is used to form a topological map by taking some risky areas as nodes, identifying critical paths and bottlenecks, and formulating corresponding resource allocation strategies.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program. It is characterized in that When the processor executes the computer program, the steps of the power project risk prediction method based on data mining described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, It is characterized in that When the computer program is executed by a processor, the steps of the power project risk prediction method based on data mining described in any one of claims 1 to 7 are implemented.

Images