Power industry software supplier evaluation system based on principal component analysis

Abstract

The invention belongs to the technical field of software supplier evaluation, and discloses a power industry software supplier evaluation system based on principal component analysis, which comprises a data acquisition module, an index definition module, a principal component analysis module, an evaluation calculation module and a visual display module, according to the power industry software supplier evaluation system based on principal component analysis, evaluation data can be automatically processed, and a visual analysis result of the comprehensive capability of the supplier can be output. Through comparing the supplier ranking generated by the system with a manual evaluation result, the accuracy and rationality of the model are evaluated. If the correlation between the model result and the manual evaluation result is significant, verifying the validity of the model; and if not, optimizing an index selection method, a weight distribution method or a principal component extraction method. Besides, in the verification process, the robustness of the horizontal type can be tested through a leave-one verification or cross verification method, and the applicability and the generalization performance of the method on different data sets are ensured.

Description

Technical Field

[0001] This invention belongs to the field of software vendor evaluation technology, and in particular relates to a software vendor evaluation system for the power industry based on principal component analysis. Background Technology

[0002] Software vendors in the power industry primarily refer to companies that focus on providing IT solutions and software services to the power sector. The software offered by these vendors typically covers multiple aspects of power engineering projects, including planning, management, monitoring, and operation, aiming to improve the operational efficiency and management level of power companies. However, current evaluation methods for power industry software vendors rely on manual assessments and cannot automatically process evaluation data to output intuitive analytical results demonstrating the vendors' comprehensive capabilities.

[0003] Based on the above analysis, the problems and shortcomings of the existing technology are as follows: Current software vendor evaluations in the power industry rely on manual assessments and cannot automatically process evaluation data or output intuitive analytical results on the vendors' overall capabilities. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention provides a power industry software supplier evaluation system based on principal component analysis.

[0005] This invention is implemented as follows: A power industry software vendor evaluation system based on principal component analysis includes: Data acquisition module, indicator definition module, principal component analysis module, evaluation calculation module, and visualization module; The data acquisition module, connected to the indicator definition module and the visualization module, is used to obtain relevant supplier data from historical supplier evaluation records, procurement contracts, and project delivery data sources in the power industry. The indicator definition module, connected to the data acquisition module, evaluation calculation module, and visualization display module, is used to include 12 evaluation indicators, including technology reserves, enterprise qualifications, quality commitment, financial strength, and software applicability. The principal component analysis module, connected to the indicator definition module and the evaluation calculation module, is used to standardize indicator data, reduce dimensionality using principal component analysis, and extract a few key principal components to characterize the supplier's comprehensive capabilities. The evaluation calculation module, connected to the indicator definition module, principal component analysis module, and visualization module, is used to calculate the supplier's comprehensive score based on the principal component score. The visualization module, connected to the data acquisition module, indicator definition module, and evaluation calculation module, is used to display evaluation results and comparative analysis of various suppliers in the form of radar charts and bar charts.

[0006] Furthermore, the data acquisition module: 1) Data collection: Use database APIs or ETL tools to import supplier data from the purchasing system or Excel files into the database; The data fields include 12 indicators such as technology reserves, enterprise qualifications, quality commitments, and basic information about suppliers; 2) Data cleaning: Handling missing values ​​in the data: ensuring data integrity by imputing missing records through the mean, interpolation, or deletion; Check for outliers in the data and identify and handle them using box plots or Z-score methods. 3) Data standardization: The indicators are normalized using the following formula:

[0007] Where X is the original value, μ is the mean, σ is the standard deviation, and X' is the standardized value.

[0008] Furthermore, the principal component analysis module: 1) Construct the covariance matrix: Calculate the covariance matrix of each indicator using the normalized data:

[0009] 2) Calculate eigenvalues ​​and eigenvectors: Use linear algebra methods or call Python's numpy.lina1g.eig function to calculate eigenvalues ​​and eigenvectors; 3) Select principal components: Based on cumulative contribution rate ( Select principal components to achieve a cumulative contribution rate of over 85%; 4) Calculate the principal component scores: Calculate the principal component scores based on the selected feature vectors and the original data:

[0010] Where X represents standardized data, and Vj represents the eigenvector of the j-th principal component.

[0011] Furthermore, the evaluation calculation module: 1) Calculation of overall score: The formula for the overall score is:

[0012] in, Principal component scores, The proportion of eigenvalues ​​to the total number of eigenvalues; 2) Supplier ranking; 3) Generate an evaluation report.

[0013] 5. The power industry software supplier evaluation system based on principal component analysis as described in claim 4, characterized in that the supplier ranking is as follows: Sort all suppliers by their overall scores in descending order to generate a supplier ranking list.

[0014] Furthermore, the generation of the evaluation report: Summarize the evaluation results, including the overall score, ranking, and scores of each indicator for each supplier, and export them as PDF or Excel files.

[0015] Another objective of this invention is to provide a method for evaluating software vendors in the power industry based on principal component analysis, comprising: Step 1: Obtain relevant supplier data from historical supplier evaluation records, procurement contracts, and project delivery data sources in the power industry through the data acquisition module; Step 2 involves defining 12 evaluation indicators, including technology reserves, enterprise qualifications, quality commitment, financial strength, and software applicability, through the indicator definition module. Step 3: Standardize the indicator data using the principal component analysis module, reduce the dimensionality using principal component analysis, and extract a few key principal components to characterize the supplier's overall capabilities. Step 4: Calculate the supplier's overall score based on the principal component scores using the evaluation calculation module; Step 5: Use the visualization module to display the evaluation results and comparative analysis of each supplier in the form of radar charts and bar charts.

[0016] Another object of the present invention is to provide a computer device including a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the power industry software vendor evaluation method based on principal component analysis.

[0017] Another object of the present invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of the power industry software vendor evaluation method based on principal component analysis.

[0018] Another objective of this invention is to provide an information data processing terminal for implementing the power industry software supplier evaluation system based on principal component analysis.

[0019] Based on the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solution to be protected by this invention are as follows: First, this invention provides a power industry software supplier evaluation system based on principal component analysis, which can automatically process evaluation data and output intuitive analysis results of supplier comprehensive capabilities. The accuracy and rationality of the model are evaluated by comparing the system-generated supplier rankings with manual evaluation results. If the model results are significantly correlated with the manual evaluation results, the effectiveness of the model is verified; if they are inconsistent, the selection of indicators, weight allocation, or principal component extraction methods can be optimized. Furthermore, during the verification process, leave-one-out validation or cross-validation methods can be used to test the robustness of the model, ensuring its applicability and generalizability on different datasets. Attached Figure Description

[0020] Figure 1 This is a block diagram of a power industry software supplier evaluation system based on principal component analysis provided in an embodiment of the present invention.

[0021] Figure 2 This is a flowchart of the data acquisition module method provided in an embodiment of the present invention.

[0022] Figure 3 This is a flowchart of the principal component analysis module method provided in an embodiment of the present invention.

[0023] Figure 4 This is a flowchart of a power industry software supplier evaluation method based on principal component analysis provided in an embodiment of the present invention.

[0024] Figure 1 The module consists of: 1. Data acquisition module; 2. Indicator definition module; 3. Principal component analysis module; 4. Evaluation and calculation module; and 5. Visualization module. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0026] like Figure 1 As shown in the figure, an embodiment of the present invention provides a power industry software supplier evaluation system based on principal component analysis, comprising: Data acquisition module 1, indicator definition module 2, principal component analysis module 3, evaluation calculation module 4, visualization display module 5; Data acquisition module 1, connected to indicator definition module 2 and visualization module 5, is used to obtain relevant supplier data from historical supplier evaluation records, procurement contracts and project delivery data sources in the power industry. The indicator definition module 2 is connected to the data acquisition module 1, the evaluation calculation module 4, and the visualization display module 5. It is used to include 12 evaluation indicators, including technology reserves, enterprise qualifications, quality commitment, financial strength, and software applicability. Principal component analysis module 3, connected to index definition module 2 and evaluation calculation module 4, is used to standardize index data, reduce dimensionality using principal component analysis, and extract a few key principal components to characterize the supplier's comprehensive capabilities. The evaluation calculation module 4, connected to the indicator definition module 2, the principal component analysis module 3, and the visualization module 5, is used to calculate the supplier's comprehensive score based on the principal component score. The visualization module 5, connected to the data acquisition module 1, the indicator definition module 2, and the evaluation calculation module 4, is used to display the evaluation results and comparative analysis of each supplier in the form of radar charts and bar charts.

[0027] The first step of the system is completed by the data acquisition module, which is responsible for obtaining relevant supplier data from multiple data sources. This data includes historical supplier evaluation records, procurement contract information, and project delivery results. The module collects this data through automated interfaces and integrates it to form a standardized dataset. The accuracy and comprehensiveness of the data provide a reliable foundation for subsequent evaluation processes and ensure the scientific validity of the evaluation results.

[0028] Next, the indicator definition module constructs 12 evaluation indicators based on the specific needs of the power industry, including technology reserves, enterprise qualifications, quality commitments, financial strength, and software applicability. These indicators cover key dimensions of supplier capabilities. The module maps the collected data to each indicator and standardizes all indicator data to eliminate unit differences and dimensional effects, so that subsequent principal component analysis can accurately extract core features.

[0029] The principal component analysis (PCA) module performs dimensionality reduction on the standardized indicator data, using PCA to extract a few key principal components to characterize the supplier's overall capabilities. For example, 12 indicators may be extracted into 2-3 principal components, which can represent the supplier's comprehensive characteristics such as technical capabilities and financial strength. Subsequently, the evaluation calculation module calculates the supplier's overall score based on the principal component scores and assigns weights to generate the final evaluation score. This stage simplifies complex multidimensional data into easily understandable scoring results through scientific statistical analysis methods.

[0030] Finally, the evaluation results are output through a visualization module. This module uses radar charts to display each supplier's performance across different evaluation indicators and bar charts to compare suppliers' overall scores, clearly reflecting each supplier's strengths and weaknesses. Through this visualization method, users can intuitively understand the evaluation results, facilitating decision-makers to quickly select suitable software suppliers. Furthermore, the evaluation report can be exported in a standard format for further reference and use by the procurement or management departments of power companies.

[0031] In summary, through modular design and principal component analysis, this system achieves a scientific and efficient supplier evaluation process, providing strong support for software procurement in the power industry.

[0032] like Figure 2 As shown, the data acquisition module provided in this embodiment of the invention: S101, Data Acquisition: Use database APIs or ETL tools to import supplier data from the purchasing system or Excel files into the database; The data fields include 12 indicators such as technology reserves, enterprise qualifications, quality commitments, and basic information about suppliers; S102, Data Cleaning: Handling missing values ​​in the data: ensuring data integrity by imputing missing records through the mean, interpolation, or deletion; Check for outliers in the data and identify and handle them using box plots or Z-score methods. S103, Data Standardization: The indicators are normalized using the following formula:

[0033] Where X is the original value, μ is the mean, σ is the standard deviation, and X' is the standardized value.

[0034] like Figure 3 As shown, the principal component analysis module provided in this embodiment of the invention: S201, Construct the covariance matrix: Calculate the covariance matrix of each indicator using the normalized data:

[0035] S202, Calculate eigenvalues ​​and eigenvectors: Use linear algebra methods or call Python's numpy.lina1g.eig function to calculate eigenvalues ​​and eigenvectors; S203, Select principal components: Based on cumulative contribution rate ( Select principal components to achieve a cumulative contribution rate of over 85%; S204, Calculate the principal component scores: Calculate the principal component scores based on the selected feature vectors and the original data:

[0036] Where X represents standardized data, and Vj represents the eigenvector of the j-th principal component.

[0037] The evaluation calculation module provided in this embodiment of the invention: 1) Calculation of overall score: The formula for the overall score is:

[0038] in, Principal component scores, The proportion of eigenvalues ​​to the total number of eigenvalues; 2) Supplier ranking; 3) Generate an evaluation report.

[0039] Supplier ranking provided in this embodiment of the invention: Sort all suppliers by their overall scores in descending order to generate a supplier ranking list.

[0040] The generated evaluation report provided in this embodiment of the invention: Summarize the evaluation results, including the overall score, ranking, and scores of each indicator for each supplier, and export them as PDF or Excel files.

[0041] The first step in the data acquisition module is data collection (step S101). The system imports supplier data from the procurement system, supplier management system, or Excel files via database API or ETL tools. The collected data fields include 12 core evaluation indicators such as technical reserves, corporate qualifications, and quality commitments, as well as basic supplier information (such as name, registered address, and contact person). During data collection, the system ensures the integrity and accuracy of the data source, forming a raw dataset that lays the foundation for subsequent data processing.

[0042] In the data cleaning phase (step S102), the system first handles missing values ​​in the data. For missing data, the module uses various methods to repair it, such as imputation using the mean or interpolation to estimate missing values, or deleting missing records according to business needs. Subsequently, the system detects outliers in the data, using box plots or Z-scores to identify significantly deviating values ​​(such as extremely high or low scores), and adjusts or removes outliers according to set rules. Through this cleaning process, the system ensures the integrity and consistency of the data.

[0043] After data cleaning, the module standardizes the data (step S103). Since different evaluation indicators may have different dimensions and value ranges (e.g., technology reserves are expressed as a percentage, while financial strength is expressed in monetary terms), the system uses Min-Max standardization or Z-score standardization to transform the data to a uniform scale. For example, all indicator values ​​are normalized to the range [0,1] to eliminate weight bias caused by differences in dimensions. Standardized data ensures the comparability and scientific validity of different indicators in subsequent analyses.

[0044] After data collection, cleaning, and standardization, the module generates a high-quality dataset, which is then output to the indicator definition module and the principal component analysis module. The standardized data contains key supplier information and possesses consistency and standardization, providing reliable data support for subsequent principal component analysis and comprehensive scoring. Through this modular design, the data collection module achieves efficient transformation from raw data to standardized input.

[0045] In summary, the data acquisition module achieves data collection, cleaning, and standardization through step-by-step processing, providing high-quality input data for the supplier evaluation system and ensuring the accuracy and scientific rigor of subsequent evaluation work.

[0046] like Figure 4 As shown in the embodiment of the present invention, a method for evaluating software vendors in the power industry based on principal component analysis includes: S301 obtains supplier-related data from historical supplier evaluation records, procurement contracts, and project delivery data sources in the power industry through a data acquisition module; S302 includes 12 evaluation indicators through the indicator definition module, including technology reserves, enterprise qualifications, quality commitment, financial strength, and software applicability. S303 standardizes the index data through the principal component analysis module, uses principal component analysis to reduce dimensionality, and extracts a few key principal components to characterize the supplier's comprehensive capabilities. S304, calculates the supplier's overall score based on the principal component score through the evaluation calculation module; S305 uses a visualization module to display evaluation results and comparative analysis of various suppliers in the form of radar charts and bar charts.

[0047] Another object of the present invention is to provide a computer device including a memory and a processor, the memory storing a computer program that, when executed by the processor, causes the processor to perform the steps of the power industry software vendor evaluation method based on principal component analysis.

[0048] Another object of the present invention is to provide a computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of the power industry software vendor evaluation method based on principal component analysis.

[0049] Another objective of this invention is to provide an information data processing terminal for implementing the power industry software supplier evaluation system based on principal component analysis.

[0050] Specific implementation of the present invention: 1. System Architecture The system consists of the following modules: Data acquisition module: Acquires supplier-related data from sources such as historical supplier evaluation records, procurement contracts, and project delivery data in the power industry.

[0051] The indicator definition module includes 12 evaluation indicators such as technology reserves, enterprise qualifications, quality commitment, financial strength, and software applicability.

[0052] Principal Component Analysis Module: Standardizes the indicator data, uses principal component analysis to reduce dimensionality, and extracts a few key principal components to characterize the supplier's overall capabilities.

[0053] Evaluation Calculation Module: Calculates the supplier's overall score based on principal component scores.

[0054] Visualization module: Displays evaluation results and comparative analysis of various suppliers through radar charts, bar charts, and other formats.

[0055] During supplier data collection, data needs to be obtained from the power industry's procurement management system or historical records. Raw data typically includes multiple dimensions, such as technological reserves, company qualifications, and financial strength. In the cleaning phase, missing values ​​are addressed first. For missing values, methods such as mean imputation, interpolation, or direct deletion of records with severe missing values ​​can be used. Secondly, outliers are checked, for example, by identifying outliers using box plots and employing upper and lower limit truncation methods to ensure data rationality and consistency.

[0056] Since different indicators have different scales and value ranges, all indicators need to be standardized to avoid excessive influence of certain indicators on the analysis results. The standardization method uses the z-score standardization formula:

[0057] Where X is the original value, Let σ be the mean and σ be the standard deviation. These are standardized values. After standardization, the mean of all indicators is 0, and the standard deviation is 1, thereby eliminating the interference of dimensions on the analysis results and ensuring that each indicator has equal importance in the model.

[0058] Constructing the covariance matrix

[0059] In principal component analysis, the covariance matrix of the indicators must first be constructed to reflect the correlation between the indicators. The elements of the covariance matrix are defined as follows:

[0060] Where Xi and Yi are the observed values ​​of two indicators, and These are their means. The covariance matrix can describe the linear relationship between the indicators, providing a basis for subsequent principal component extraction.

[0061] Eigenvalues ​​and eigenvectors are extracted from the covariance matrix. Eigenvalues ​​reflect the variance contribution rate of each principal component, and eigenvectors define the orientation of the principal components. Eigenvalues ​​and eigenvectors are calculated using linear algebra methods and sorted in descending order of eigenvalues. Principal components with a cumulative contribution rate exceeding 85% are selected, and their scores are calculated.

[0062] in, Let X be the score of the j-th principal component, and X be the standardized data matrix. Let be the eigenvector of the j-th principal component.

[0063] The supplier's overall score is calculated based on the scores of each principal component and their corresponding weights:

[0064] in, The weights of the principal components are calculated using the following formula:

[0065] in, Let be the eigenvalue of the j-th principal component. After the comprehensive score is calculated, suppliers are sorted by score, and ranking results and evaluation reports are generated, including radar charts and tabular displays.

[0066] The accuracy and rationality of the model are evaluated by comparing the supplier rankings generated by the system with the results of human evaluation. If the model results are significantly correlated with the human evaluation results, the effectiveness of the model is verified; if they are inconsistent, the selection of indicators, weight allocation, or principal component extraction methods can be optimized. In addition, during the validation process, leave-one-out validation or cross-validation methods can be used to test the robustness of the model, ensuring its applicability and generalization on different datasets.

[0067] The specific implementation method of the present invention.

[0068] 1. Data Acquisition and Preprocessing

[0069] 1.1. Data Acquisition: Use database APIs or ETL tools (such as Talend or Apache Nifi) to import supplier data from the purchasing system or Excel files into the database.

[0070] The data fields include 12 indicators such as technology reserves, enterprise qualifications, and quality commitments, as well as basic supplier information (such as enterprise name and industry qualifications).

[0071] 1.2. Data Cleaning: Handling missing values ​​in the data: Ensure data integrity by imputing missing records through the mean, interpolation, or deletion.

[0072] Check for outliers in the data and identify and handle them using box plots or Z-score methods.

[0073] 1.3. Data Standardization: The indicators are normalized using the following formula:

[0074] Where X is the original value, μ is the mean, σ is the standard deviation, and X' is the standardized value.

[0075] 2. Principal Component Analysis

[0076] 2.1. Construct the covariance matrix: Calculate the covariance matrix of each indicator using the normalized data:

[0077] 2.2. Calculate eigenvalues ​​and eigenvectors: Use linear algebra methods or call Python's numpy.1ina1g.eig function to calculate eigenvalues ​​and eigenvectors.

[0078] 2.3. Selection of principal components: Based on cumulative contribution rate ( Select principal components to achieve a cumulative contribution rate of over 85%.

[0079] 2.4. Calculate the principal component scores: Calculate the principal component scores based on the selected feature vectors and the original data:

[0080] Where X represents standardized data, and Vj represents the eigenvector of the j-th principal component.

[0081] 3. Overall Score and Ranking

[0082] 3.1. Calculation of overall score: The formula for the overall score is:

[0083] in, Principal component scores, This represents the proportion of eigenvalues ​​to the total number of eigenvalues.

[0084] 3.2. Supplier Ranking: Sort all suppliers by their overall scores in descending order to generate a supplier ranking list.

[0085] 3.3. Generate an evaluation report: Summarize the evaluation results, including the overall score, ranking, and scores of each indicator for each supplier, and export them as PDF or Excel files.

[0086] 4. System Verification

[0087] 4.1. Comparative Verification: The system-generated supplier rankings were compared with manually evaluated rankings to verify the rationality of the principal component analysis model. 4.2 Performance Testing: Test the system's computation time and result accuracy with large-scale data (e.g., 500 suppliers). 4.3. Optimization: Adjust the data preprocessing rules or principal component weight allocation method based on the validation results to improve the robustness of the model. System operation process Users upload supplier data through the front-end display terminal.

[0088] After the data is stored in the database, the backend system automatically performs data cleaning, standardization, and principal component analysis.

[0089] The system outputs supplier rankings and displays the results visually on the front end, including a rating radar chart and a ranking table.

[0090] Users can download the evaluation report or conduct further data analysis.

[0091] Key technical indicators

[0092] Data processing time: < 10 seconds (500 suppliers).

[0093] Ranking accuracy: > 90% (compared to human evaluation results).

[0094] System stability: Supports simultaneous access by multiple users with no noticeable delay.

[0095] This implementation method integrates the statistical methods of principal component analysis with modern hardware and software environments, and can be used in practical applications for supplier evaluation in the power industry.

[0096] The system consists of the following hardware and software devices and modules: Hardware equipment: A high-performance server (CPU ≥ 8 cores, memory ≥ 16GB, storage ≥ 1TB) is used for data storage and analysis.

[0097] Network connectivity devices used for data transmission and system deployment.

[0098] Front-end display terminals (such as PCs and tablets) are used to visualize the evaluation results.

[0099] Software environment: Data storage: MySQL or PostgreSQL database, used to store supplier evaluation data.

[0100] Data analysis: Python (pandas, numpy, scikit-learn, etc.) or R language for data cleaning and principal component analysis.

[0101] Front-end presentation: Web applications based on Django, Flask, or Node.js, used for user interaction and result display.

[0102] 3. Format of deliverables

[0103] A software supplier evaluation system for the power industry based on principal component analysis can automatically process evaluation data and output intuitive analysis results of the supplier's overall capabilities.

[0104] Application Example 1: Evaluation of Software Vendors in the Power Industry

[0105] background

[0106] A power company plans to procure a business management system and needs to conduct a comprehensive evaluation of multiple candidate software vendors to ensure that the final selected vendor meets the requirements in terms of technical reserves, project delivery capabilities, and financial strength.

[0107] Implementation process

[0108] Data Acquisition and Cleaning: Relevant data on candidate suppliers were extracted from the company's historical project database, including 12 indicators (such as technology reserves, software delivery capabilities, and innovation capabilities).

[0109] Data cleaning: imputing missing values ​​(using the mean), removing outliers (detecting them using box plots), and standardizing each indicator.

[0110] Principal component analysis: Calculate the index covariance matrix.

[0111] Eigenvalues ​​and eigenvectors were extracted, and the top three principal components were selected based on the cumulative contribution rate, which reached 87%.

[0112] Calculate the score for each supplier on the three principal components.

[0113] 3. Overall Score and Ranking: According to the formula:

[0114] Calculate the overall score, where The weight of the principal component.

[0115] Suppliers are ranked based on their overall scores, generating ranking tables and radar charts to visually display the evaluation results for each supplier.

[0116] 4. Evaluation Report: Output a detailed evaluation report, including an overall score, ranking, and scores for each indicator, to support decision-making.

[0117] Application Example 2: Software Code Feature Identification and Vulnerability Analysis

[0118] background

[0119] A financial institution is conducting a security audit of its core business system and hopes to use convolutional neural network technology to automatically identify code features and generate a list of known vulnerabilities.

[0120] Implementation process

[0121] Data preprocessing: Collect the target source code and binary files, decompress them, and divide them into code segments of fixed length.

[0122] The code snippets are vectorized, mapping code elements to numerical features.

[0123] CNN Feature Extraction: Design a multi-layer convolutional neural network, inputting code snippet features.

[0124] Low-level features of the code are extracted using convolutional kernels, while high-level semantic features are captured through multiple convolutions and pooling.

[0125] Code profile generation: The extracted high-level features are clustered and matched to a database of known components to generate code profiles, including component names and version numbers.

[0126] Vulnerability Analysis: By linking code profiles with the CVE vulnerability database, known vulnerabilities in relevant components can be identified.

[0127] Output a list of vulnerabilities, including vulnerability ID, severity, and remediation recommendations.

[0128] Result verification: Test known code snippets to verify the consistency between the vulnerability list generated by the system and the results of manual analysis.

[0129] Application Example 3: Automatic Detection of Supply Chain Security Risks

[0130] A manufacturing company wants to conduct a security risk assessment of the software used in its supply chain in order to identify and prevent potential security threats in a timely manner.

[0131] Data collection: Collect component information and version numbers for all software in the supply chain.

[0132] Code recognition technology is used to generate component profiles for software components.

[0133] Attack Simulation: Based on component profiling, an attack simulation model is trained using knowledge graph technology to automatically generate vulnerability exploitation tools (PoC samples).

[0134] Simulate attacks on software components in the supply chain to test their vulnerability patching status and security protection capabilities.

[0135] Risk assessment: Generate detailed risk assessment reports for unpatched vulnerabilities, including the potential threat level, the scope of systems that may be affected, and the remediation priority.

[0136] Security recommendations: Provide remediation solutions and security protection suggestions for each vulnerability.

[0137] Regularly update component profiles and vulnerability databases to ensure the timeliness and accuracy of security assessments.

[0138] These examples demonstrate the end-to-end application from vendor evaluation to code signature identification and vulnerability analysis, which can solve practical problems in different scenarios and provide efficient and accurate technical support for the industry.

[0139] I. Specific application areas or related products of this invention

[0140] This invention has wide applications in software code security analysis, supply chain security assessment, and information system vulnerability detection. Its core technologies can provide solutions for multiple industries, including finance, manufacturing, and energy. For example, in the financial sector, it can be used for security audits of core business systems, quickly identifying code vulnerabilities and providing remediation suggestions; in the manufacturing industry, it can be applied to the security assessment of software components in the supply chain, discovering potential threats and generating protection strategies. Furthermore, the results of this invention can also serve as technical support for related products such as code security detection tools, vulnerability scanning tools, and supply chain security assessment platforms, providing comprehensive support for enterprise information security management.

[0141] II. Evidence related to the technical effects obtained by the embodiments of the present invention

[0142] Evidence: In Example 2, feature extraction and component profiling of code snippets were performed using a convolutional neural network, successfully identifying multiple high-risk vulnerabilities. These vulnerabilities were then matched against the CVE database to generate a vulnerability list. Test results show that the consistency between automated analysis and manual review is over 95%.

[0143] Evidence: In Example 3, the system utilizes code recognition and knowledge graph technologies to quickly detect vulnerabilities in multiple components of the supply chain and tests the remediation status of vulnerabilities through attack simulations. The risk assessment report generated by the system covers potential threat levels and remediation priorities, with assessment efficiency significantly higher than traditional manual methods.

[0144] Evidence: In both embodiments, the system not only identifies vulnerabilities but also provides specific remediation solutions (such as upgrading component versions or adjusting code configurations). According to testing, 90% of the remediation suggestions can be directly adopted and effectively resolve security issues.

[0145] Evidence: By regularly updating the vulnerability database and component profiles, the system can quickly respond to new security threats, ensuring the real-time nature and accuracy of assessment results. This was fully verified in the supply chain risk assessment of Example 3.

[0146] In summary, this invention, through modular design and efficient analysis technology, has been successfully applied to multiple industry scenarios, providing users with accurate and reliable security detection and risk assessment services.

[0147] It should be noted that embodiments of the present invention can be implemented in hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the above-described devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuitry such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of the above-described hardware circuitry and software, such as firmware.

[0148] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A power industry software vendor evaluation system based on principal component analysis, characterized in that, include: Data acquisition module, indicator definition module, principal component analysis module, evaluation calculation module, and visualization module; The data acquisition module, connected to the indicator definition module and the visualization module, is used to obtain relevant supplier data from historical supplier evaluation records, procurement contracts, and project delivery data sources in the power industry. The indicator definition module, connected to the data acquisition module, evaluation calculation module, and visualization display module, is used to include 12 evaluation indicators, including technology reserves, enterprise qualifications, quality commitment, financial strength, and software applicability. The principal component analysis module, connected to the indicator definition module and the evaluation calculation module, is used to standardize indicator data, reduce dimensionality using principal component analysis, and extract a few key principal components to characterize the supplier's comprehensive capabilities. The evaluation calculation module, connected to the indicator definition module, principal component analysis module, and visualization module, is used to calculate the supplier's comprehensive score based on the principal component score. The visualization module, connected to the data acquisition module, indicator definition module, and evaluation calculation module, is used to display evaluation results and comparative analysis of various suppliers in the form of radar charts and bar charts.

2. The power industry software supplier evaluation system based on principal component analysis as described in claim 1, characterized in that, The data acquisition module: 1) Data collection: Use database APIs or ETL tools to import supplier data from the purchasing system or Excel files into the database; The data fields include 12 indicators such as technology reserves, enterprise qualifications, quality commitments, and basic information about suppliers; 2) Data cleaning: Handling missing values ​​in the data: ensuring data integrity by imputing missing records through the mean, interpolation, or deletion; Check for outliers in the data and identify and handle them using box plots or Z-score methods. 3) Data standardization: The indicators are normalized using the following formula: Where X is the original value, μ is the mean, σ is the standard deviation, and X' is the standardized value.

3. The power industry software supplier evaluation system based on principal component analysis as described in claim 1, characterized in that, The principal component analysis module: 1) Construct the covariance matrix: Calculate the covariance matrix of each indicator using the normalized data: 2) Calculate eigenvalues ​​and eigenvectors: Use linear algebra methods or call Python's numpy.lina1g.eig function to calculate eigenvalues ​​and eigenvectors; 3) Select principal components: Based on cumulative contribution rate ( Select principal components to achieve a cumulative contribution rate of over 85%; 4) Calculate the principal component scores: Calculate the principal component scores based on the selected feature vectors and the original data: Where X represents standardized data, and Vj represents the eigenvector of the j-th principal component.

4. The power industry software supplier evaluation system based on principal component analysis as described in claim 1, characterized in that, The evaluation calculation module: 1) Calculation of overall score: The formula for the overall score is: in, Principal component scores, The proportion of eigenvalues ​​to the total number of eigenvalues; 2) Supplier ranking; 3) Generate an evaluation report.

5. The power industry software supplier evaluation system based on principal component analysis as described in claim 4, characterized in that, The supplier ranking: Sort all suppliers by their overall scores in descending order to generate a supplier ranking list.

6. The power industry software supplier evaluation system based on principal component analysis as described in claim 4, characterized in that, The generated evaluation report: Summarize the evaluation results, including the overall score, ranking, and scores of each indicator for each supplier, and export them as PDF or Excel files.

7. A method for evaluating power industry software suppliers based on principal component analysis, implementing the power industry software supplier evaluation system based on principal component analysis as described in any one of claims 1-6, characterized in that, The power industry software vendor evaluation method based on principal component analysis includes: Step 1: Obtain relevant supplier data from historical supplier evaluation records, procurement contracts, and project delivery data sources in the power industry through the data acquisition module; Step 2 involves defining 12 evaluation indicators, including technology reserves, enterprise qualifications, quality commitment, financial strength, and software applicability, through the indicator definition module. Step 3: Standardize the indicator data using the principal component analysis module, reduce the dimensionality using principal component analysis, and extract a few key principal components to characterize the supplier's overall capabilities. Step 4: Calculate the supplier's overall score based on the principal component scores using the evaluation calculation module; Step 5: Use the visualization module to display the evaluation results and comparative analysis of each supplier in the form of radar charts and bar charts.

8. A computer device, characterized in that, The computer device includes a memory and a processor. The memory stores a computer program that, when executed by the processor, causes the processor to perform the steps of the power industry software supplier evaluation method based on principal component analysis as described in claim 7.

9. A computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of the power industry software vendor evaluation method based on principal component analysis as described in claim 7.

10. An information data processing terminal, characterized in that, The information data processing terminal is used to implement the power industry software supplier evaluation system based on principal component analysis as described in any one of claims 1-6.

Images