An intelligent evaluation method and system for power project investment based on artificial intelligence algorithm

Abstract

The present invention discloses an intelligent evaluation method and system for power project investment based on artificial intelligence algorithm. The method includes reviewing the feasibility of the project in the feasibility study stage of the power project, including evaluation of technical feasibility, economic feasibility and other aspects; conducting intelligent post-evaluation of the investment effect of the project in the investment stage of the power project, using artificial intelligence algorithm to analyze and evaluate various indicators of the investment project, including evaluation of investment return rate, asset-liability ratio and other aspects; and conducting intelligent operation and maintenance evaluation of the wind turbine in the operation and maintenance stage of the power project, analyzing and evaluating the operation data of the wind turbine, including evaluation of the performance, operation status, fault diagnosis and other aspects of the wind turbine. The system includes an intelligent post-evaluation subsystem for power project investment, a feasibility study report review subsystem, and an intelligent operation and maintenance evaluation subsystem for the wind turbine. The present invention can comprehensively and accurately evaluate the feasibility, investment effect and operation status of the power project.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an intelligent evaluation method and system for power project investment based on artificial intelligence algorithm. Background Art

[0002] The post-evaluation system of power projects refers to a system that evaluates and analyzes the implementation effects and results of power projects after the completion of power projects. The system can evaluate the achievement of project goals, cost-effectiveness, environmental impact, social benefits, etc. by collecting, organizing and analyzing relevant data and information of the project, and provide evaluation results and suggestions so that project managers and decision makers can make reasonable decisions and optimization measures. The rating of power construction power project enterprises is the process of evaluating and scoring the comprehensive strength and credit status of power project enterprises. Ratings are usually based on a series of indicators and standards, including financial data, operating conditions, market position, management capabilities and industry prospects. The purpose of rating is to provide investors, financial institutions, suppliers and other stakeholders with an assessment of the credit risk and value of power project enterprises so that they can make decisions and take corresponding measures. Power project enterprise ratings are usually expressed in letters or symbols, such as AAA, AA, A, BBB and BB. Different rating agencies may adopt different rating standards and rating systems, but generally speaking, a high rating means that the credit of the power project enterprise is good and has a lower risk of default, while a low rating means that the credit of the power project enterprise is poor and the risk is higher. Power project enterprise ratings can help investors and financial institutions better understand the risks and potential of power project enterprises so as to make investment decisions and provide financial services such as loans or credit. At the same time, ratings can also help power project enterprises improve their transparency and creditworthiness and enhance their market competitiveness.

[0003] Prior art 1, application number: 202310052766.3 discloses a personnel and enterprise rating method and system based on multidimensional data, which obtains multidimensional data of personnel or enterprises, divides the multidimensional data of personnel or enterprises into internal data and external data and performs preprocessing; uses Flume and kafka to store the preprocessed internal data and external data of personnel or enterprises in HDFS files respectively; obtains the data of the person or enterprise to be predicted from the HDFS file and inputs it into the trained random forest model, and outputs the rating of the person or enterprise to be predicted. Although the multidimensional data of personnel or enterprises are processed separately according to different data types and then uniformly stored in the HDFS file, it is convenient to call the multidimensional data of personnel or enterprises later. However, the rating accuracy is poor and the processing efficiency is low, which affects the efficiency of the rating.

[0004] Prior art 2, application number: 202310927394.4 discloses a data analysis method and server for enterprise rating, the method comprising: after the server obtains the enterprise sample data, it pre-processes the enterprise sample data to obtain the selected data, screens the elements in the selected data according to their importance, obtains the characteristic data, classifies the sample enterprises, obtains the sub-sample enterprise set, determines the classification ranking of the sample enterprises in the sub-sample set and the corresponding ranking score according to the weight coefficient of each element, constructs a rating model, inputs the enterprise information of the enterprise to be rated into the rating model, and obtains the rating result of the enterprise to be rated. Although through this method, users such as investors or financiers can directly see the evaluation results of the enterprise among enterprises with similar competitiveness, can accurately see the development potential of the enterprise, and understand the potential competitiveness of the enterprise in detail. However, the indicators and factors of the sample data are limited, which reduces the accuracy of the enterprise rating evaluation results.

[0005] Prior art three, application number: 202310112196.2 discloses a method for classifying characteristics of evaluation of a technology enterprise rating model, which determines the complexity level of the database by the amount of credit data in the database, classifies the credit data of the database of the first complexity level according to the characteristic keywords pre-stored in the characteristic database, and stores the credit data in the sub-database to which it belongs, determines the sub-database that can be used to train the classification model according to the amount of credit data stored in each sub-database, inputs the credit data in the sub-database that can be used to train the classification model into the classification model, obtains the evaluation result according to the classification result, determines whether the classification model is qualified according to the qualified rate of the evaluation result, and adjusts the preset impact factor comparison threshold in the classification model when the classification model is judged to be unqualified to obtain a qualified classification model to classify the credit data, which improves the accuracy and scientificity of the classification of technology enterprise credit data. However, the real-time and dynamic performance are poor.

[0006] At present, the existing technologies 1, 2 and 3 have the problem that the real-time and dynamic nature of sample data is poor, which reduces the accuracy and efficiency of power project enterprise rating. Therefore, the present invention provides a method and system for intelligent evaluation of power project investment based on artificial intelligence algorithm. Summary of the invention

[0007] The main purpose of the present invention is to provide an intelligent evaluation method and system for power project investment based on artificial intelligence algorithm, so as to solve the problem that the real-time and dynamic performance of sample data in the prior art is poor, which reduces the accuracy and efficiency of power project enterprise rating.

[0008] To achieve the above object, the present invention provides the following technical solutions:

[0009] An intelligent evaluation method for power project investment based on artificial intelligence algorithm, the intelligent evaluation method for power project investment based on artificial intelligence algorithm comprising:

[0010] During the feasibility study phase of a power project, the feasibility of the project is reviewed, including the evaluation of technical feasibility, economic feasibility and market prospects;

[0011] During the investment phase of power projects, the investment effects of the projects are intelligently evaluated, and various indicators of investment projects are analyzed and evaluated using artificial intelligence algorithms, including the evaluation of return on investment, debt-to-asset ratio, and cash flow;

[0012] During the operation and maintenance phase of the power project, the intelligent operation and maintenance of the wind turbine is evaluated, and the operating data of the wind turbine is analyzed and evaluated, including the evaluation of the wind turbine's performance, operating status, and fault diagnosis.

[0013] As a further improvement of the present invention, the relevant data includes financial data, market data and industry data.

[0014] As a further improvement of the present invention, the process of establishing a rating model using a neural network includes:

[0015] Divide the data set containing historical data into a training set and a validation set, with a ratio of 70%-30%; use the training set to train the rating model, and establish the model by learning the patterns and rules of the data; in the neural network, select the multi-layer perceptron and activation function, and initialize the parameters of the rating model;

[0016] During the training process, the training set is used to optimize the weights of the neural network; the back propagation algorithm and stochastic gradient descent are used to update the neural network parameters to reduce the loss function of the evaluation model; through multiple iterative training, the rating model is gradually adjusted and optimized;

[0017] The trained rating model is verified and evaluated using the validation set, and the performance of the rating model is evaluated using accuracy.

[0018] As a further improvement of the present invention, the process of gradually adjusting and optimizing the rating model includes:

[0019] The input data of the training set is forward propagated through the neural network to obtain the output of the neural network; the output of the neural network is compared with the true label and the value of the loss function is calculated;

[0020] Starting from the output layer, the chain rule is used to calculate the gradient of the loss function for each parameter. Through back propagation, the gradient is transferred from the output layer to the input layer layer by layer to calculate the gradient value of each parameter. The gradient descent algorithm is used to update the parameters of the neural network according to the calculated parameter gradient.

[0021] Repeat until a predetermined number of iterations or convergence condition of the loss function is reached; in each iteration, different training samples or sample batches are used to update the parameters to increase the randomness and generalization ability of training.

[0022] As a further improvement of the present invention, the rating model is verified and evaluated, including:

[0023] The input data of the validation set is forward propagated through the trained rating model to obtain the output of the rating model. According to the output of the rating model, classification or prediction operations are performed to obtain the prediction results of the rating model for the validation set samples.

[0024] Compare the prediction results of the rating model with the true labels of the validation set and calculate the accuracy of the model on the validation set;

[0025] Based on the calculated accuracy, the performance of the rating model on the validation set is evaluated; the higher the accuracy, the better the performance of the rating model in predicting ratings; based on the evaluation results of the validation set, the rating model is adjusted and optimized.

[0026] As a further improvement of the present invention, the user views the rating results and related indicators of the power project enterprise through an interface or a report.

[0027] As a further improvement of the present invention, the process of presenting the rating results to the user in a visual form includes:

[0028] Provide an arrangement area and condition configuration area for rating prediction results, and add elements related to rating prediction results through the arrangement area; present the rating prediction results to users in the form of a visual bar chart, which contains the rating results and related indicators of the power project enterprise;

[0029] The visual rating results are embedded into the web page or application interface; users can input the power project enterprise information or select the power project enterprise of interest through the interactive operation of the interface to view its rating results and related indicators;

[0030] Generate a report containing power project enterprise rating results and related indicators based on user needs.

[0031] As a further improvement of the present invention, the basic information, filtering conditions, aggregation conditions and sorting rules of the elements are configured in the conditional configuration area to obtain a visual configuration of the bar chart association analysis.

[0032] As a further improvement of the present invention, the report includes a detailed description of the rating results, an explanation and analysis of the rating indicators, and a comparison with other power project enterprises or industry average levels.

[0033] To achieve the above object, the present invention also provides the following technical solutions:

[0034] An intelligent evaluation system for power project investment based on artificial intelligence algorithm, which is applied to the intelligent evaluation method for power project investment based on artificial intelligence algorithm, and the intelligent evaluation system for power project investment based on artificial intelligence algorithm includes:

[0035] The power project investment intelligent post-evaluation subsystem is used to intelligently evaluate the investment effect of power projects through artificial intelligence algorithms, providing a comprehensive and accurate evaluation of investment effects;

[0036] Feasibility study report review subsystem, used to review the feasibility study report of the power project;

[0037] The wind turbine intelligent operation and maintenance evaluation subsystem is used to evaluate the intelligent operation and maintenance of the wind turbine. By analyzing and evaluating the operating data of the wind turbine, it provides information on the fan performance, operating status, and fault diagnosis.

[0038] The present invention reviews and evaluates the feasibility of power projects to ensure that the projects are technically, economically and marketably feasible; helps enterprises avoid investing in ineffective or risky projects and improves the accuracy and reliability of investment decisions; uses artificial intelligence algorithms to analyze and evaluate various indicators of investment projects, including evaluations of return on investment, asset-liability ratio, cash flow, etc.; helps enterprises evaluate the risks and returns of investment projects, optimize investment portfolios, and improve investment results and economic benefits. Evaluates the intelligent operation and maintenance of wind turbines, analyzes and evaluates the operating data of wind turbines, including evaluations of wind turbine performance, operating status, fault diagnosis, etc.; helps enterprises realize intelligent operation and maintenance management of wind turbines, improves the operating efficiency and reliability of wind turbines, and reduces operation and maintenance costs and risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a schematic diagram of the steps of an embodiment of the method for intelligent evaluation of power project investment based on artificial intelligence algorithm of the present invention;

[0040] Figure 2 This is a schematic diagram of a process of performing intelligent post-evaluation of a project's investment effect according to an embodiment of the intelligent evaluation method for power project investment based on an artificial intelligence algorithm of the present invention;

[0041] Figure 3A schematic diagram of a process for obtaining relevant data of a power project enterprise according to an embodiment of the intelligent evaluation method for power project investment based on an artificial intelligence algorithm of the present invention;

[0042] Figure 4 A schematic diagram of a process of establishing a rating model using a neural network in one embodiment of an intelligent evaluation method for power project investment based on an artificial intelligence algorithm of the present invention;

[0043] Figure 5 A schematic diagram of a process of presenting rating results to users in a visual form according to an embodiment of the intelligent evaluation method for power project investment based on artificial intelligence algorithm of the present invention;

[0044] Figure 6 This is a functional module diagram of an embodiment of an intelligent evaluation system for power project investment based on artificial intelligence algorithm of the present invention;

[0045] Figure 7 It is a structural schematic diagram of an embodiment of an electronic device of the present invention;

[0046] Figure 8 It is a schematic structural diagram of an embodiment of the storage medium of the present invention. DETAILED DESCRIPTION

[0047] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0048] The terms "first", "second" and "third" in the present invention are only used for descriptive purposes and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Therefore, the features defined as "first", "second" and "third" can explicitly or implicitly include at least one of the features. In the description of the present invention, the meaning of "multiple" is at least two, such as two, three, etc., unless otherwise clearly and specifically defined. All directional indications (such as up, down, left, right, front, back...) in the embodiments of the present invention are only used to explain the relative position relationship, movement, etc. between the components under a certain specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indication also changes accordingly. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally also includes steps or units that are not listed, or optionally also includes other steps or units inherent to these processes, methods, products or devices.

[0049] Reference to an "embodiment" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiment may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0050] like Figure 1 As shown, this embodiment provides an embodiment of an intelligent evaluation method for power project investment based on an artificial intelligence algorithm. In this embodiment, the intelligent evaluation method for power project investment based on an artificial intelligence algorithm specifically includes the following steps:

[0051] Step S1: During the feasibility study phase of a power project, the feasibility of the project is reviewed, including evaluation of technical feasibility, economic feasibility and market prospects;

[0052] Step S2: During the investment phase of the power project, the investment effect of the project is intelligently evaluated, and the various indicators of the investment project are analyzed and evaluated using artificial intelligence algorithms, including evaluation of return on investment, debt-to-asset ratio, cash flow, etc.;

[0053] Step S3: During the operation and maintenance phase of the power project, the intelligent operation and maintenance of the wind turbine is evaluated, and the operation data of the wind turbine is analyzed and evaluated, including evaluation of the performance, operation status, fault diagnosis, etc. of the wind turbine.

[0054] Preferably, step S1 of this embodiment reviews and evaluates the feasibility of the power project to ensure that the project is feasible in terms of technology, economy and market; helps enterprises avoid investing in ineffective or risky projects and improves the accuracy and reliability of investment decisions. Step S2 uses artificial intelligence algorithms to analyze and evaluate various indicators of investment projects, including evaluations of return on investment, debt-to-asset ratio, cash flow, etc.; helps enterprises evaluate the risks and returns of investment projects, optimize investment portfolios, and improve investment effects and economic benefits. Step S3 evaluates the intelligent operation and maintenance of wind turbines, analyzes and evaluates the operating data of wind turbines, including evaluations of wind turbine performance, operating status, fault diagnosis, etc.; helps enterprises realize intelligent operation and maintenance management of wind turbines, improves the operating efficiency and reliability of wind turbines, and reduces operation and maintenance costs and risks.

[0055] In summary, the application of the enterprise rating method based on artificial intelligence algorithm in different stages of this embodiment can effectively improve the decision-making accuracy and economic benefits of enterprises, reduce risks and costs, and enhance the competitiveness and sustainable development capabilities of enterprises. Through the evaluation and review of these three subsystems, the feasibility, investment effect and operation status of power projects can be comprehensively and accurately evaluated, providing investors with a basis for decision support and operation and maintenance optimization.

[0056] Furthermore, if Figure 2 As shown, the intelligent post-evaluation of the investment effect of the project in step S2 specifically includes the following steps:

[0057] Step S21: Obtain relevant data of the power project enterprise, the relevant data including financial data, market data and industry data, etc.; pre-process the collected relevant data; extract target features from the pre-processed relevant data according to the evaluation indicators and dimensions of the power project enterprise rating;

[0058] Step S22: Use a neural network to establish a rating model, and use historical data to optimize and adjust the parameters of the rating model;

[0059] Step S23: The rating model makes rating predictions based on the relevant data of the new power project enterprise, and presents the rating results to the user in a visual form; the user views the rating results and related indicators of the power project enterprise through the interface or report.

[0060] Preferably, step S21 of this embodiment provides necessary input data for the rating model by obtaining financial data, market data and industry data of the power project enterprise; improves the quality and availability of the data by cleaning, deduplication and normalization of the collected data; extracts features related to the evaluation indicators and dimensions of the rating from the preprocessed data to provide effective input for the rating model. Step S22 uses a neural network to establish a rating model: by using a neural network algorithm, a model that can learn and predict the rating of power project enterprises is established; parameter optimization and adjustment: the parameters of the rating model are optimized and adjusted through historical data to improve the accuracy and performance of the model. Step S23 Rating prediction: the rating model performs rating prediction based on the new power project enterprise-related data to predict the comprehensive strength and credit status of the power project enterprise; visual presentation: the rating results are presented to the user in a visual form to provide intuitive rating results and related indicators.

[0061] Through the above steps, this embodiment can achieve the following direct technical effects: data preparation and preprocessing can provide high-quality and consistent data, providing reliable input for the rating model; feature extraction can extract important features related to the rating from the data, improving the accuracy and explanatory power of the rating model; the rating model established using a neural network has high accuracy and predictive power, and can make accurate rating predictions for power project companies; the visual presentation of rating results can enable users to intuitively understand the rating results and related indicators of the power project companies, providing better decision support.

[0062] In this embodiment, by using a neural network model and a large amount of data for training and optimization, the rating model can more accurately predict the strength and credit status of the power project enterprise; by extracting relevant features from the data, the rating model can consider multiple indicators and factors, thereby providing a more comprehensive evaluation result; the automation and optimization of data preprocessing and model training can improve the efficiency of the rating process and save time and resources; the rating model can make real-time rating predictions based on the latest power project enterprise data, and promptly reflect the changes and risks of the power project enterprise; by visually presenting the rating results and related indicators, the rating system can automatically generate reports and charts, and provide user-friendly rating interpretation and analysis.

[0063] Furthermore, if Figure 3 As shown, the process of obtaining relevant data of the power project enterprise in step S21 specifically includes the following steps:

[0064] Step S211: Acquire device configuration information storing relevant data, obtain storage paths and data names of relevant data from the device configuration information, parse the device configuration information, and classify relevant data according to the data names;

[0065] Step S212: different data sets of related data are obtained according to the classification, and the parts with intersections in the different data sets are identified and extracted as the recognition data sets. After preprocessing the recognition data sets of different related data, each item of data in the recognition data set is converted into a feature vector and uploaded to the feature recognition model;

[0066] Step S213: comparing the feature vectors in the recognition data sets of different related data, and treating the inconsistent text data in the recognition data of each related data as abnormal data, and feeding the abnormal data back to the recognition data set for elimination, thereby obtaining a data set without abnormal data;

[0067] Step S214: According to the evaluation index and dimension of the rating, perform feature dimensionality reduction operation on the data set to extract target features; perform label encoding processing on the extracted target features.

[0068] Preferably, step S211 of this embodiment obtains the device configuration information for storing relevant data: obtains the storage path and data name of the relevant data from the device configuration information to provide a data source for subsequent operations; parses the device configuration information: parses the device configuration information and converts it into a format that can be understood and processed; classifies the relevant data: classifies the relevant data according to the data name to distinguish different types of data. Step S212 identifies the extraction and preprocessing of the data set: obtains the identification data set by identifying and extracting the parts with intersections in different data sets, and performs preprocessing operations on it to ensure the accuracy and consistency of the data; feature vector conversion and upload: converts each data in the identification data set into a feature vector and uploads it to the feature recognition model for subsequent data comparison and analysis. Step S213 identifies and removes abnormal data: identifies the inconsistent parts as abnormal data by comparing the feature vectors in the identification data sets of different relevant data, and feeds them back to the identification data set for elimination to improve the quality and accuracy of the data. Step S214: Feature dimension reduction and target feature extraction: Based on the evaluation indicators and dimensions of the rating, perform feature dimension reduction on the data set to extract target features related to the rating for use in training and prediction of the rating model; Label encoding processing: Perform label encoding processing on the extracted target features so that the rating model can understand and process these features.

[0069] Through the above steps, this embodiment can realize a series of operations from obtaining relevant data of power project enterprises to extracting target features; the direct technical effects of these operations include data classification and preprocessing, abnormal data identification and elimination, target feature extraction and label encoding processing to improve data quality and accuracy, and provide accurate and useful feature input for the rating model.

[0070] Furthermore, the process of extracting target features in step S214 specifically includes the following steps:

[0071] Step S2141: Collect historical data and establish a training data set containing category labels; for each category, calculate the mean vector of its samples to represent the center position of the category in the feature space; calculate the scatter matrix within each category to measure the degree of dispersion of samples within the category; calculate the scatter matrix between categories to measure the differences between different categories;

[0072] Step S2142: Based on the intra-class scatter matrix and the inter-class scatter matrix, the solution of the generalized eigenvalue problem is calculated, and the eigenvector corresponds to the eigenvalue and represents the direction of the low-dimensional subspace after the LDA transformation; the eigenvalues ​​are sorted in descending order, and the first k eigenvectors are selected, where k is the number of feature dimensions that you want to retain;

[0073] Step S2143: The selected eigenvectors are organized into a transformation matrix by columns, where the matrix is ​​a projection matrix of the LDA transformation; the original data set is multiplied by the transformation matrix to map it into a low-dimensional subspace.

[0074] Step S2144: extract target features related to the evaluation indicators and dimensions of the rating from the features after dimensionality reduction; perform label encoding on the extracted target features to convert them into a form that can be understood and processed by the rating model; label encoding can use methods such as one-hot encoding and label encoding.

[0075] Preferably, the evaluation indicators and dimensions in this embodiment are standards and elements used to measure the comprehensive strength and credit status of power project enterprises; the specific content may vary according to the specific needs and business scenarios of the rating, for example: financial status includes: balance sheet indicators: total assets, total liabilities, net assets and debt-to-asset ratio, etc.; income statement indicators: operating income, net profit, gross profit margin and net profit margin, etc.; cash flow statement indicators: cash flow from operating activities, cash flow from investment activities and cash flow from financing activities, etc. Market position includes: market share: the share of power project enterprises in a specific market; brand awareness: the recognition and reputation of the power project enterprise brand; customer satisfaction: the degree of customer satisfaction with the products and services of the power project enterprise. Management capabilities include: business strategy: the business strategy and planning formulated and implemented by the power project enterprise; leadership team: the ability and experience of the senior management team of the power project enterprise; innovation capability: the R&D investment and innovation results of the power project enterprise. Market risks include: industry competitiveness: the competitive situation and position of the industry in which the power project enterprise is located; policy environment: the impact of government policies on the operation of the power project enterprise; macroeconomic environment: the impact of macroeconomic factors on the operation of the power project enterprise.

[0076] In this embodiment, the category mean vector, for category c, its mean vector is marked as m c , the category mean vector expression is obtained by calculating the mean of the feature vectors of all sample data within the category:

[0077]

[0078] Among them, n c represents the number of samples in category c, x ci Represents the feature vector of the i-th sample in category c;

[0079] The intra-class scatter matrix, for category c, is denoted by S c , the expression is obtained by calculating the covariance matrix of the difference between the samples within the category and the category mean eigenvector:

[0080]

[0081] The inter-class scatter matrix is ​​denoted as S B , the expression is obtained by calculating the covariance matrix of the difference between the mean vectors of different categories:

[0082]

[0083] Among them, C represents the total number of categories, n c represents the number of samples in category c, and m represents the overall mean vector of all samples.

[0084] In this embodiment, step S2141: collect historical data and establish a training data set to provide the necessary data basis for subsequent feature extraction and model training; calculate the mean vector of the category, and obtain the central position of each category in the feature space, providing a benchmark for the subsequent category difference analysis; calculate the intra-class scatter matrix and the inter-class scatter matrix to measure the discrete degree of samples within the category and the difference between different categories, providing a basis for feature extraction and dimensionality reduction operations. Step S2142: solve the generalized eigenvalue problem, obtain a set of eigenvectors and corresponding eigenvalues, these eigenvectors represent the direction of the low-dimensional subspace after LDA transformation, and have the greatest category difference; sort in descending order according to the size of the eigenvalue, and select the first k eigenvectors, where k is the number of feature dimensions you want to retain, so that you can select the most discriminative features for subsequent feature extraction and dimensionality reduction operations. Step S2143: Generate a transformation matrix, and organize the selected feature vectors into a transformation matrix by columns. This matrix is ​​the projection matrix of the LDA transformation, which is used to map the original data to a low-dimensional subspace; multiply the original data set by the transformation matrix to achieve dimensionality reduction of the data, map the data to a low-dimensional feature space, and retain the most discriminative features. Step S2144: Extract target features, and select target features related to the evaluation indicators and dimensions of the rating from the reduced-dimensional features. These features can better reflect the strength and credit status of the power project company; perform label encoding on the extracted target features and convert them into a form that can be understood and processed by the rating model to facilitate the training and prediction of the rating model.

[0085] Through the above steps, this embodiment can extract target features related to rating indicators and dimensions from the original data and convert them into a calculable and comparable form, providing a basis for the training and prediction of the rating model. At the same time, the dimensionality reduction operation can reduce redundant features and noise, improve the modeling effect and the model's explanatory power.

[0086] Furthermore, if Figure 4 As shown, the process of using a neural network to establish a rating model in step S22 specifically includes the following steps:

[0087] Step S221: Divide the data set containing historical data into a training set and a validation set, with a ratio of 70%-30%; use the training set to train the rating model, and establish the model by learning the patterns and rules of the data; in the neural network, select a multilayer perceptron and an activation function, and initialize the parameters of the rating model;

[0088] Step S222: During the training process, the weights of the neural network are optimized using the training set; the neural network parameters are updated using the back propagation algorithm and stochastic gradient descent to reduce the loss function of the evaluation model; and the rating model is gradually adjusted and optimized through multiple iterations of training;

[0089] Step S223: Use the validation set to validate and evaluate the trained rating model, and use accuracy to evaluate the performance of the rating model.

[0090] Preferably, step S221 of this embodiment: divide the data set into a training set and a validation set to facilitate the evaluation of the performance and generalization ability of the model; use the training set to train the rating model, and establish the model by learning the patterns and rules of the data; select a multi-layer perceptron as the structure of the neural network, and the activation function is used to introduce nonlinear factors. Step S222: Use the training set to optimize the weights of the neural network, and update the neural network parameters through optimizers such as back propagation algorithm and stochastic gradient descent; during the optimization process, the model is adjusted and optimized by learning the patterns and rules of the data to reduce the loss function of the evaluation model and improve the prediction ability of the model. Step S223: Use the validation set to verify and evaluate the trained rating model, and evaluate the performance and prediction accuracy of the model; through evaluation indicators such as accuracy, the prediction ability and generalization ability of the rating model on the validation set can be measured.

[0091] Through the above steps, this embodiment can divide the data set into a training set and a validation set, and use the training set to train and optimize the rating model; using the multi-layer perceptron structure and activation function of the neural network, a rating model with good predictive ability can be established; through the evaluation of the validation set, the performance of the model can be evaluated and verified to further optimize and improve the accuracy and generalization ability of the rating model.

[0092] Furthermore, the process of establishing a model by learning the patterns and rules of data in step S221 specifically includes the following steps:

[0093] Step S2211: combining the target analysis results of the data in the training set with the Bernoulli distribution and the target analysis result detection neural network, alternately optimizing and performing data enhancement training operations, and obtaining a self-inspired learned data enhancement operation sequence;

[0094] Step S2212: The original data and the data enhancement operation sequence in the training set are transmitted to the feature pyramid network for target network data analysis, and attribute labels of different target data are output;

[0095] Step S2213: Using the target data processed by the feature pyramid network as input and the target data marked with attributes as true annotations, the detection neural network in the evaluation model is trained to obtain a trained evaluation model of different target data with attribute labels.

[0096] Preferably, step S2211 of this embodiment: by combining the data target analysis results in the training set with the Bernoulli distribution, a self-inspired data enhancement operation sequence can be obtained; the operation sequence can help the training model better understand the patterns and laws of the data, thereby improving the accuracy and robustness of the model; through the self-learning data enhancement operation sequence, the model can optimize the characteristics of the data in a targeted manner, thereby improving the training effect. Step S2212: the feature pyramid network receives the original data and the data enhancement operation sequence as input, and outputs attribute tags with different target data; the feature pyramid network can help the model extract multi-level features of the data, thereby more accurately identifying and marking different target data; through the processing of the feature pyramid network, the model's ability to understand and express data can be improved. Step S2213: by taking the target data processed by the feature pyramid network as input, and the target data marked with attributes as the real annotation, the detection network in the evaluation model is trained; the evaluation model obtained by training can be used to detect and identify different target data with attribute tags; by training the evaluation model, the model's ability to recognize and classify target data can be further improved.

[0097] This embodiment optimizes feature extraction and target recognition through self-learning data enhancement operation sequences, thereby improving the accuracy, robustness, and generalization ability of the model; through self-learning and optimization, the model is able to better understand and apply data, thereby providing more accurate, comprehensive, and personalized answers and suggestions.

[0098] Furthermore, the process of gradually adjusting and optimizing the rating model in step S222 specifically includes the following steps:

[0099] Step S2221: forward propagate the input data of the training set through the neural network to obtain the output result of the neural network; compare the output result of the neural network with the true label and calculate the value of the loss function;

[0100]

[0101] Among them, D(x) represents the feature vector of the output result, D(y) represents the feature vector of the true label, ||D(x)-D(y)|| 2The Euclidean distance between the feature vector of the output result and the feature vector of the true label, r represents the upper limit of the loss function of the output result of the neural network, B(x,y)=1 means that the output result of the neural network matches the true label, (x,y)=0 means that the output result of the neural network does not match the true label.

[0102] Step S2222: Starting from the output layer, the chain rule is used to calculate the gradient of the loss function for each parameter; the gradient is transferred from the output layer to the input layer layer by layer through back propagation, and the gradient value of each parameter is calculated; the gradient descent algorithm is used to update the parameters of the neural network according to the calculated parameter gradient;

[0103] Step S2223: Repeat steps 2221 to 2222 until a predetermined number of iterations or the convergence condition of the loss function is reached; in each iteration, different training samples or sample batches are used to update parameters to increase the randomness and generalization ability of training.

[0104] Preferably, step S2221 of this embodiment: by forward propagating the input data of the training set through the neural network, the output result of the neural network is obtained, and it is compared with the true label to calculate the value of the loss function; the difference between the prediction result of the model and the true label can be measured to evaluate the performance and accuracy of the model. Step S2222: by using the chain rule to calculate the gradient of the loss function for each parameter, and by back propagating the gradient from the output layer to the input layer layer by layer, the gradient value of each parameter is calculated; then, the parameters of the neural network are updated according to the calculated parameter gradient using the gradient descent algorithm; the adjustment and optimization of the parameters can be guided according to the gradient of the loss function, so that the model can better fit the training data. Step S2223: by repeatedly executing steps S2221 and S2222 until a predetermined number of iterations or convergence conditions of the loss function are reached; in each iteration, different training samples or sample batches are used to update the parameters to increase the randomness and generalization ability of the training; the parameters of the model can be continuously adjusted and optimized to gradually adapt to the patterns and laws of the training data, and improve the performance and prediction ability of the model.

[0105] This embodiment adjusts and optimizes the rating model by calculating the value of the loss function and the parameter gradient, as well as continuous parameter updates; through continuous iteration and optimization, the model is able to better fit the training data and improve the accuracy, robustness, and generalization ability of the model; by gradually adjusting and optimizing the parameters of the model, the rating model can better predict and classify data and provide more accurate and useful rating results.

[0106] Furthermore, the verification and evaluation of the rating model in step S223 specifically includes the following steps:

[0107] Step S2231: forward propagating the input data of the validation set through the trained rating model to obtain the output result of the rating model, and performing classification or prediction operations based on the output result of the rating model to obtain the prediction result of the rating model for the validation set samples;

[0108] Step S2232: Compare the prediction results of the rating model with the true labels of the validation set, and calculate the accuracy of the model on the validation set;

[0109] Accuracy calculation formula: Accuracy = (number of correctly predicted samples) / (total number of samples)

[0110] Step S2233: Based on the calculated accuracy, evaluate the performance of the rating model on the validation set; the higher the accuracy, the better the performance of the rating model in predicting ratings; based on the evaluation results of the validation set, the rating model can be adjusted and optimized. For example, the hyperparameters, network structure or loss function of the rating model can be adjusted to improve the performance of the rating model.

[0111] Preferably, step S2231 of this embodiment: forward propagate the input data of the validation set through the trained rating model to obtain the output result of the rating model, and perform classification or prediction operations based on the output result to obtain the prediction result of the rating model for the validation set sample; the rating model can be used to predict the rating of the data in the validation set to obtain the prediction result of the model in practical application. Step S2232: compare the prediction result of the rating model with the true label of the validation set, and calculate the accuracy of the model on the validation set; the accuracy calculation formula is Accuracy = (number of correctly predicted samples) / (total number of samples); the accuracy is an indicator to measure the accuracy of the model prediction, and the performance of the rating model on the validation set can be evaluated. Step S2233: evaluate the performance of the rating model on the validation set based on the calculated accuracy; the higher the accuracy, the better the performance of the rating model in predicting ratings; based on the evaluation results of the validation set, the rating model can be adjusted and optimized, such as adjusting hyperparameters, network structure or loss function, etc., to improve the performance and accuracy of the rating model.

[0112] To summarize, this embodiment uses the rating model to predict and calculate the accuracy of the validation set to evaluate the performance and accuracy of the rating model in practical applications; through the evaluation results of the validation set, we can understand the prediction accuracy of the rating model, judge the pros and cons of the rating model, and make corresponding adjustments and optimizations to improve the performance and reliability of the rating model in practical applications; through verification and evaluation, we can ensure the stability and generalization ability of the rating model on different data, and improve the application value and reliability of the rating model in practical applications.

[0113] Furthermore, if Figure 5As shown, the process of presenting the rating results to the user in a visual form in step S23 specifically includes the following steps:

[0114] Step S231: providing an arrangement area and a condition configuration area for rating prediction results, adding elements related to the rating prediction results through the arrangement area; configuring basic information, filtering conditions, aggregation conditions and sorting rules of the elements through the condition configuration area to obtain a visual configuration of the bar chart association analysis; presenting the rating prediction results to the user in the form of a visual bar chart, wherein the bar chart includes the rating results and related indicators of the power project enterprise for display;

[0115] Step S232: embedding the visualized rating results into a web page or application program interface; the user inputs the power project enterprise information or selects the power project enterprise of interest through interactive operations on the interface to view its rating results and related indicators;

[0116] Step S233: Generate a report containing the rating results and related indicators of the power project enterprise according to user needs; the report includes a detailed description of the rating results, an explanation and analysis of the rating indicators, and a comparison with other power project enterprises or industry average levels.

[0117] Preferably, step S231 of this embodiment: realizes the visual configuration of the rating prediction results, including the selection, screening, aggregation and sorting of the rating results and related indicators of the power project enterprises to be displayed in the bar chart; the display form of the rating results can be customized according to the needs and preferences of the user, and the rating elements and related indicators of the power project enterprises of interest can be selected to perform screening, aggregation and sorting operations; the user can understand the rating of the power project enterprise more intuitively according to his or her own focus and quickly obtain key information. Step S232 is embedded in the interface, so that the user can easily query and browse, realize the visual display and interactive access of the rating results; the rating results are directly presented to the user in the form of a bar chart, embedded in the user's commonly used interface, such as a web page or application program interface; it can improve the convenience and operation efficiency of the user, allowing the user to access the rating results anytime and anywhere, and facilitate the user to understand the rating of the power project enterprise in real time. Step S233: Through the generation of the report, the user can gain an in-depth understanding of the meaning of the rating results and indicators, and conduct more comprehensive and detailed analysis and decision-making; generate a detailed rating result report based on user needs; provide a more in-depth and comprehensive rating result analysis to help users better understand the prediction results of the rating model and make more accurate decisions; at the same time, the generation of the report also helps users record and share the rating results, making it convenient to share rating information with others.

[0118] In summary, this embodiment implements the visual configuration, interface embedding, and report generation of rating results, so that users can easily view and analyze the rating results and related indicators of power project enterprises, and provide a more intuitive and interactive way to understand and apply the prediction results of the rating model. Through the above steps, the rating model can make rating predictions based on the new power project enterprise-related data, and present the rating results to the user in a visual form. Users can conveniently view the rating results and related indicators of power project enterprises through interfaces or reports, helping them make decisions or understand the risk status of power project enterprises. At the same time, by displaying the rating results in a visual form, users' understanding and trust in the rating model can be improved.

[0119] like Figure 6 As shown, this embodiment also provides an embodiment of an intelligent evaluation system for power project investment based on artificial intelligence algorithm. In this embodiment, the intelligent evaluation system for power project investment based on artificial intelligence algorithm is applied to the intelligent evaluation method for power project investment based on artificial intelligence algorithm in the above-mentioned embodiment. The intelligent evaluation system for power project investment based on artificial intelligence algorithm includes an intelligent post-evaluation subsystem 1 for power project investment, a feasibility study report review subsystem 2, and a wind turbine intelligent operation and maintenance evaluation subsystem 3 which are electrically connected in sequence.

[0120] Among them, the power project investment intelligent post-evaluation subsystem 1 is used to conduct intelligent post-evaluation of the investment effect of the power project, and use artificial intelligence algorithms to analyze and evaluate various indicators of the investment project, including investment return rate, asset-liability ratio, cash flow, etc.; through evaluation, it can provide investors with a comprehensive and accurate investment effect evaluation, help make reasonable investment decisions, and optimize the operation and management of the project. The feasibility study report review subsystem 2 is used to review the feasibility study report of the power project, analyze and evaluate the various contents in the report, including the technical feasibility, economic feasibility, market prospects, etc. of the project; through review, it can provide investors with an objective and accurate evaluation to help them make investment decisions. The wind turbine intelligent operation and maintenance evaluation subsystem 3 is used to evaluate the intelligent operation and maintenance of the wind turbine, analyze and evaluate the operation data of the wind turbine, including the performance, operating status, fault diagnosis, etc. of the wind turbine; through evaluation, it can provide reliability, efficiency and safety information of the wind turbine operation, help operation and maintenance personnel to timely discover and solve problems, and improve the operation effect of the wind turbine.

[0121] Preferably, the power project investment intelligent post-evaluation subsystem 1 of this embodiment uses an artificial intelligence algorithm to intelligently evaluate the investment effect of the power project, provide a comprehensive and accurate investment effect evaluation, help investors make reasonable investment decisions, and optimize the operation and management of the project; help investors better understand the project's rate of return, asset-liability ratio, cash flow and other indicators, and improve the accuracy and efficiency of investment decisions. The feasibility study report review subsystem 2 provides an objective and accurate evaluation by reviewing the feasibility study report of the power project, helping investors understand the technical feasibility, economic feasibility, market prospects and other aspects of the project; provides decision-making basis for investors, helps them evaluate the feasibility and potential risks of the project, and thus makes wise investment decisions. The wind turbine intelligent operation and maintenance evaluation subsystem 3 evaluates the intelligent operation and maintenance of the wind turbine, and provides information on wind turbine performance, operating status, fault diagnosis and other aspects by analyzing and evaluating the operation data of the wind turbine; improves the efficiency and reliability of wind turbine operation and maintenance, helps operation and maintenance personnel to discover and solve problems in a timely manner, improves the operation effect of the wind turbine, and reduces maintenance costs.

[0122] In summary, this embodiment provides accurate and objective evaluation and decision support by applying artificial intelligence algorithms, helping investors and operation and maintenance personnel to achieve better technical effects and economic benefits in power project investment and wind turbine operation and maintenance.

[0123] Furthermore, the power project investment intelligent post-evaluation subsystem 1 specifically includes:

[0124] The data preprocessing module is used to obtain relevant data of power project enterprises, including financial data, market data, industry data, etc.; preprocess the collected relevant data; extract target features from the preprocessed relevant data according to the evaluation indicators and dimensions of the power project enterprise rating;

[0125] Model building module, which is used to build rating models using neural networks and optimize and adjust rating model parameters using historical data;

[0126] The rating result output module is used for the rating model to make rating predictions based on the relevant data of the new power project enterprise, and present the rating results to the user in a visual form; the user views the rating results and related indicators of the power project enterprise through the interface or report.

[0127] Preferably, this embodiment provides necessary input data for the rating model by obtaining financial data, market data, industry data, etc. of the power project enterprise; improves the quality and availability of the data by cleaning, deduplication, normalization, etc. of the collected data; extracts features related to the evaluation indicators and dimensions of the rating from the preprocessed data to provide effective input for the rating model. Use neural networks to establish a rating model: By using a neural network algorithm, establish a model that can learn and predict the rating of power project enterprises; Parameter optimization and adjustment: Optimize and adjust the parameters of the rating model through historical data to improve the accuracy and performance of the model. Rating prediction: The rating model makes rating predictions based on new power project enterprise-related data to predict the comprehensive strength and credit status of the power project enterprise; Visual presentation: The rating results are presented to the user in a visual form to provide intuitive rating results and related indicators.

[0128] This embodiment can achieve the following direct technical effects: data preparation and preprocessing can provide high-quality and consistent data, providing reliable input for the rating model; feature extraction can extract important features related to the rating from the data, improving the accuracy and explanatory power of the rating model; the rating model established using a neural network has high accuracy and predictive power, and can make accurate rating predictions for power project companies; the visual presentation of rating results can enable users to intuitively understand the rating results and related indicators of the power project company, providing better decision support.

[0129] In this embodiment, by using a neural network model and a large amount of data for training and optimization, the rating model can more accurately predict the strength and credit status of the power project enterprise; by extracting relevant features from the data, the rating model can consider multiple indicators and factors, thereby providing a more comprehensive evaluation result; the automation and optimization of data preprocessing and model training can improve the efficiency of the rating process and save time and resources; the rating model can make real-time rating predictions based on the latest power project enterprise data, and promptly reflect the changes and risks of the power project enterprise; by visually presenting the rating results and related indicators, the rating system can automatically generate reports and charts, and provide user-friendly rating interpretation and analysis.

[0130] Furthermore, the data preprocessing module specifically includes:

[0131] The information configuration submodule is used to obtain the device configuration information storing the relevant data, obtain the storage path and data name of the relevant data from the device configuration information, parse the device configuration information, and classify the relevant data according to the data name;

[0132] The data recognition submodule is used to obtain different data sets of related data according to classification, identify and extract the parts with intersections in different data sets as recognition data sets, pre-process the recognition data sets of different related data respectively, convert each data in the recognition data set into a feature vector and upload it to the feature recognition model;

[0133] The vector comparison submodule is used to compare the feature vectors in the recognition data sets of different related data. The inconsistent parts of the text data in the recognition data of each related data are regarded as abnormal data. The abnormal data are fed back to the recognition data set for elimination to obtain a data set without abnormal data.

[0134] According to the evaluation indicators and dimensions of the rating, the data set is subjected to feature dimensionality reduction operations to extract the target features; the extracted target features are subjected to label encoding processing.

[0135] Preferably, the present embodiment obtains the device configuration information for storing relevant data: obtains the storage path and data name of the relevant data from the device configuration information to provide a data source for subsequent operations; parses the device configuration information: parses the device configuration information and converts it into a format that can be understood and processed; classifies the relevant data: classifies the relevant data according to the data name to distinguish different types of data. Extraction and preprocessing of identification data sets: by identifying and extracting the parts with intersections in different data sets, an identification data set is obtained, and preprocessing operations are performed on it to ensure the accuracy and consistency of the data; feature vector conversion and uploading: each item of data in the identification data set is converted into a feature vector and uploaded to the feature recognition model for subsequent data comparison and analysis. Abnormal data identification and elimination: by comparing the feature vectors in the identification data sets of different relevant data, the inconsistent parts are identified as abnormal data, and they are fed back to the identification data set for elimination to improve the quality and accuracy of the data. Feature dimensionality reduction and target feature extraction: Based on the rating evaluation indicators and dimensions, perform feature dimensionality reduction on the data set to extract target features related to the rating for use in rating model training and prediction; Label encoding processing: Perform label encoding processing on the extracted target features so that the rating model can understand and process these features.

[0136] This embodiment can realize a series of operations from acquiring relevant data of power project enterprises to extracting target features; the direct technical effects of these operations include data classification and preprocessing, abnormal data identification and elimination, target feature extraction and label encoding processing to improve data quality and accuracy, and provide accurate and useful feature input for the rating model.

[0137] Furthermore, the model building module specifically includes:

[0138] The data set partitioning submodule is used to partition the data set containing historical data into a training set and a validation set, with a partition ratio of 70%-30%; the rating model is trained using the training set, and the model is established by learning the patterns and rules of the data; in the neural network, a multi-layer perceptron and activation function are selected, and the parameters of the rating model are initialized;

[0139] The weight optimization submodule is used to optimize the weights of the neural network using the training set during the training process; the back propagation algorithm and stochastic gradient descent are used to update the neural network parameters to reduce the loss function of the evaluation model; and the rating model is gradually adjusted and optimized through multiple iterative training;

[0140] The model evaluation submodule is used to verify and evaluate the trained rating model using the validation set, and the performance of the rating model is evaluated by accuracy.

[0141] Preferably, this embodiment divides the data set into a training set and a validation set to facilitate the evaluation of the performance and generalization ability of the model; the training set is used to train the rating model, and the model is established by learning the patterns and regularities of the data; a multilayer perceptron is selected as the structure of the neural network, and the activation function is used to introduce nonlinear factors. The weights of the neural network are optimized using the training set, and the neural network parameters are updated by optimizers such as the back propagation algorithm and stochastic gradient descent; during the optimization process, the model is adjusted and optimized by learning the patterns and regularities of the data to reduce the loss function of the evaluation model and improve the predictive ability of the model. The trained rating model is verified and evaluated using the validation set to evaluate the performance and prediction accuracy of the model; the prediction ability and generalization ability of the rating model on the validation set can be measured through evaluation indicators such as accuracy.

[0142] This embodiment can divide the data set into a training set and a validation set, and use the training set to train and optimize the rating model; using the multi-layer perceptron structure and activation function of the neural network, a rating model with good predictive ability can be established; through the evaluation of the validation set, the performance of the model can be evaluated and verified to further optimize and improve the accuracy and generalization ability of the rating model.

[0143] Furthermore, the rating result output module specifically includes:

[0144] The visual configuration submodule is used to provide the arrangement area and condition configuration area of ​​the rating prediction results, and add the elements related to the rating prediction results through the arrangement area; configure the basic information, filtering conditions, aggregation conditions and sorting rules of the elements through the condition configuration area to obtain the visual configuration of the bar chart association analysis; present the rating prediction results to the user in the form of a visual bar chart, which contains the rating results and related indicators of the power project enterprises for display;

[0145] The interface setting submodule is used to embed the visual rating results into a web page or application program interface; users can input power project enterprise information or select power project enterprises of interest through interactive operations on the interface to view their rating results and related indicators;

[0146] The report generation submodule is used to generate a report containing the rating results and related indicators of the power project enterprise according to user needs; the report includes a detailed description of the rating results, an explanation and analysis of the rating indicators, and a comparison with other power project enterprises or the industry average level.

[0147] Preferably, this embodiment realizes the visualization configuration of the rating prediction results, including the selection, screening, aggregation and sorting of the rating results and related indicators of the power project enterprises to be displayed in the bar chart; the display form of the rating results can be customized according to the needs and preferences of the user, and the rating elements and related indicators of the power project enterprises of interest can be selected to perform screening, aggregation and sorting operations; the user can understand the rating of the power project enterprise more intuitively according to his or her focus and quickly obtain key information. Through interface embedding, users can easily query and browse, realize the visualization display and interactive access of the rating results; the rating results are directly presented to the user in the form of a bar chart, embedded in the user's commonly used interface, such as a web page or application program interface; the user's convenience and operating efficiency can be improved, allowing the user to access the rating results anytime and anywhere, and it is convenient for the user to understand the rating of the power project enterprise in real time. Through the generation of reports, users can gain a deeper understanding of the meaning of rating results and indicators, and conduct more comprehensive and detailed analysis and decision-making; generate detailed rating result reports based on user needs; provide a more in-depth and comprehensive analysis of rating results, help users better understand the prediction results of the rating model, and make more accurate decisions; at the same time, the generation of reports also helps users record and share rating results, making it easier to share rating information with others.

[0148] In summary, this embodiment implements the visual configuration, interface embedding, and report generation of rating results, so that users can easily view and analyze the rating results and related indicators of power project enterprises, and provide a more intuitive and interactive way to understand and apply the prediction results of the rating model. Through the above steps, the rating model can make rating predictions based on the new power project enterprise-related data, and present the rating results to the user in a visual form. Users can conveniently view the rating results and related indicators of power project enterprises through interfaces or reports, helping them make decisions or understand the risk status of power project enterprises. At the same time, by displaying the rating results in a visual form, users' understanding and trust in the rating model can be improved.

[0149] Furthermore, the feasibility study report review subsystem 2 specifically includes:

[0150] The wind pressure and snow pressure value database design module is used to design and maintain a database to store relevant data on wind pressure and snow pressure, including wind pressure and snow pressure values ​​for various regions and building types; the database can provide accurate data support for the review of feasibility study reports;

[0151] Remote area classification database design module is used to design and maintain a database to store classification information of remote areas, including geographical location, climate conditions, transportation and other factors; the database can provide remote area-related data reference for feasibility study report review;

[0152] The earthquake peak acceleration adjustment coefficient database design module is used to design and maintain a database to store the relevant data of the earthquake peak acceleration adjustment coefficient; according to different regions and building types, it can provide accurate earthquake parameter data for feasibility study report review;

[0153] The database design module for labor budget unit price standards for construction and installation is used to design and maintain a database to store labor budget unit price standards for construction and installation projects; including labor cost standards for various building materials and engineering projects, which can provide accurate labor budget data for feasibility study report review;

[0154] The earthquake intensity parameter database design module is used to design and maintain a database to store relevant data of earthquake intensity parameters, including the impact of different earthquake intensity levels on buildings and the corresponding parameter values, which can provide accurate earthquake parameter data for feasibility study report review;

[0155] The solar energy formula maintenance function development module is used to develop the maintenance function of solar energy formulas, including solar radiation calculation formulas, photovoltaic power generation calculation formulas, etc.; it can easily update and maintain solar energy related formulas to provide accurate solar energy data for feasibility study report review;

[0156] The electrical formula maintenance function development module is used to develop the maintenance function of electrical formulas, including current calculation formulas, power calculation formulas, etc.; it can easily update and maintain electrical-related formulas to provide accurate electrical data for feasibility study report review;

[0157] The power generation formula maintenance function development module is used to develop the maintenance function of the power generation formula, including the wind power generation calculation formula, solar power generation calculation formula, etc.; it can easily update and maintain the formula related to power generation, and provide accurate power generation data for the feasibility study report review;

[0158] The system test case writing module is used to write system test cases, including functional testing and performance testing of the feasibility study report review subsystem; it can ensure the stability and reliability of the system and provide accurate evaluation and test results for the feasibility study report review.

[0159] Preferably, the wind pressure and snow pressure value database design module of this embodiment can ensure that the influence of wind pressure and snow pressure is taken into account when designing and evaluating power project investment, thereby improving the feasibility and safety of the project. The remote area classification database design module provides remote area-related data reference for feasibility study report review, helping to evaluate the feasibility and risk of power project investment in remote areas. The earthquake peak acceleration adjustment coefficient database design module can provide accurate earthquake parameter data for feasibility study report review, which can ensure that earthquake risks are taken into account when designing and evaluating power project investment, thereby improving the feasibility and earthquake resistance of the project. The building installation labor budget unit price standard database design module can provide accurate labor budget data for feasibility study report review, which helps to evaluate the cost-effectiveness and economic feasibility of power project investment. The earthquake intensity parameter database design module can provide accurate earthquake parameter data for feasibility study report review, which helps to evaluate the earthquake risk and safety of power project investment. The solar energy formula maintenance function development module facilitates the update and maintenance of solar energy-related formulas, provides accurate solar energy data for feasibility study report review, and helps to evaluate the potential and feasibility of solar power generation projects. The electrical formula maintenance function development module facilitates the updating and maintenance of electrical-related formulas, provides accurate electrical data for feasibility study report review, and helps evaluate the electrical system design and efficiency of power project investment. The power generation formula maintenance function development module facilitates the updating and maintenance of power generation-related formulas, provides accurate power generation data for feasibility study report review, and helps evaluate the power generation capacity and benefits of power project investment. The system test case writing module writes system test cases to ensure the functionality and performance of the feasibility study report review subsystem, helps verify the stability and reliability of the system, and ensures the accuracy of the evaluation and test results.

[0160] In summary, this embodiment provides accurate data and parameter support for feasibility study report review, helping to evaluate the feasibility, safety, economy and reliability of power project investment, which can improve the scientificity and accuracy of decision-making, reduce project risks, and promote sustainable development.

[0161] Furthermore, the wind turbine intelligent operation and maintenance evaluation subsystem 3 specifically includes:

[0162] Data acquisition and processing module, used to collect the fan operation data in real time, including wind speed, rotation speed, temperature, humidity and other parameters;

[0163] The feature extraction and analysis module is used to extract features from the collected data and extract features related to the performance and operating status of the fan; for example, the power curve, vibration frequency and other features of the fan can be extracted. Then, the features are analyzed to identify the operating status and performance indicators of the fan;

[0164] The fault diagnosis and prediction module is used to diagnose the faults of the fan operation data. By learning from historical data and known fault cases, it can identify the potential faults of the fan and predict the probability and impact of the faults.

[0165] The operation and maintenance optimization suggestion module is used to give targeted operation and maintenance optimization suggestions based on the analysis and evaluation of wind turbine data. For example, it can make suggestions such as adjusting the parameter settings of the wind turbine and replacing faulty parts to improve the operation effect and life of the wind turbine.

[0166] Preferably, the data acquisition and processing module of this embodiment can obtain accurate information on the operation status of the fan by collecting the operation data of the fan in real time, and provide a data basis for subsequent feature extraction and analysis. The feature extraction and analysis module can identify features related to the performance and operation status of the fan by extracting and analyzing the collected data. These features can be used to evaluate the operation status and performance indicators of the fan, such as power curve, vibration frequency, etc.; by analyzing the features, potential problems or abnormal conditions can be quickly identified to provide support for fault diagnosis and prediction. Fault diagnosis and prediction module: Based on the learning of historical data and known fault cases, the operation data of the fan is diagnosed; by identifying potential faults and predicting the probability and impact of the faults, it can help operation and maintenance personnel to find and solve problems in time and improve the reliability and safety of the fan. The operation and maintenance optimization suggestion module gives targeted operation and maintenance optimization suggestions based on the analysis and evaluation of the fan data; by proposing suggestions such as adjusting parameter settings and replacing faulty parts, the operation effect of the fan can be improved, the life of the fan can be extended, the maintenance cost can be reduced, and higher energy utilization can be achieved.

[0167] To summarize, this embodiment improves the operating effect and reliability of the fan: through the analysis and evaluation of the fan operation data, it can timely discover potential faults and problems, provide operation and maintenance optimization suggestions, and help operation and maintenance personnel take timely measures to reduce the occurrence of faults and improve the reliability and stability of the fan; increase the operating efficiency of the fan: through the evaluation of the performance and operating status of the fan, it can find the reasons for the low operating efficiency, and put forward optimization suggestions to help improve the energy utilization efficiency of the fan and reduce operating costs; enhance the safety of the fan: through the fault diagnosis and prediction module, it can identify the potential faults of the fan, and predict the probability of occurrence and impact of the fault, helping operation and maintenance personnel to take timely measures to ensure the safe operation of the fan; reduce operation and maintenance costs: through the intelligent operation and maintenance evaluation subsystem, it can realize real-time monitoring and evaluation of the fan operation status, timely discover problems, reduce maintenance and repair costs, and improve operation and maintenance efficiency.

[0168] like Figure 7 As shown, this embodiment provides an embodiment of an electronic device. In this embodiment, the electronic device 4 includes a processor 41 and a memory 42 coupled to the processor 41.

[0169] The memory 42 stores program instructions for implementing any of the above-mentioned embodiments of the intelligent evaluation method for power project investment based on artificial intelligence algorithm.

[0170] The processor 41 is used to execute program instructions stored in the memory 42 to perform intelligent evaluation of power project investment based on artificial intelligence algorithms.

[0171] The processor 41 may also be referred to as a CPU (Central Processing Unit). The processor 41 may be an integrated circuit chip having signal processing capabilities. The processor 41 may also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0172] Further, Figure 8The schematic diagram of the structure of the storage medium of an embodiment of the present application is that the storage medium 5 of the embodiment of the present application stores program instructions 51 that can implement all the above methods, wherein the program instructions 51 can be stored in the above storage medium in the form of a software product, including several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, or terminal devices such as a computer, a server, a mobile phone, and a tablet.

[0173] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0174] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above integrated unit may be implemented in the form of hardware or in the form of software functional units. The above is only an implementation mode of the present invention, and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention specification and drawings, or directly or indirectly used in other related technical fields, is also included in the patent protection scope of the present invention.

[0175] The specific implementation methods of the invention are described in detail above, but they are only examples, and the invention is not limited to the specific implementation methods described above. For those skilled in the art, any equivalent modification or substitution of the invention is also within the scope of the invention, therefore, the equalization, modification, improvement, etc. made without departing from the spirit and principle of the invention should be included in the scope of the invention.

Claims

1. An intelligent evaluation method for power project investment based on artificial intelligence algorithm, characterized in that: The intelligent evaluation method for power project investment based on artificial intelligence algorithm includes: During the feasibility study phase of a power project, the feasibility of the project is reviewed, including the evaluation of technical feasibility, economic feasibility and market prospects; During the investment phase of power projects, intelligent post-evaluation of the investment effects of the projects is carried out, and artificial intelligence algorithms are used to analyze and evaluate various indicators of investment projects, including evaluation of return on investment, debt-to-asset ratio, and cash flow; During the operation and maintenance phase of the power project, the intelligent operation and maintenance of the wind turbine is evaluated, and the operating data of the wind turbine is analyzed and evaluated, including the evaluation of the wind turbine's performance, operating status, and fault diagnosis.

2. The intelligent evaluation method for power project investment based on artificial intelligence algorithm according to claim 1 is characterized in that: Conduct intelligent post-evaluation of the project's investment effect, including: Obtain relevant data of power project enterprises and pre-process the collected relevant data; extract target features from the pre-processed relevant data according to the evaluation indicators and dimensions of power project enterprise rating; the relevant data include financial data, market data and industry data; Use neural networks to build rating models and use historical data to optimize and adjust rating model parameters; The rating model makes rating predictions based on the relevant data of new power project companies and presents the rating results to users in a visual form.

3. The intelligent evaluation method for power project investment based on artificial intelligence algorithm according to claim 2 is characterized in that: The process of building a rating model using a neural network includes: Divide the data set containing historical data into a training set and a validation set, with a ratio of 70%-30%; use the training set to train the rating model, and establish the model by learning the patterns and rules of the data; in the neural network, select the multi-layer perceptron and activation function, and initialize the parameters of the rating model; During the training process, the training set is used to optimize the weights of the neural network; the back propagation algorithm and stochastic gradient descent are used to update the neural network parameters to reduce the loss function of the evaluation model; through multiple iterative training, the rating model is gradually adjusted and optimized; The trained rating model is verified and evaluated using the validation set, and the performance of the rating model is evaluated using accuracy.

4. The intelligent evaluation method for power project investment based on artificial intelligence algorithm according to claim 3 is characterized in that: The process of gradually adjusting and optimizing the rating model includes: The input data of the training set is forward propagated through the neural network to obtain the output of the neural network; the output of the neural network is compared with the true label and the value of the loss function is calculated; Starting from the output layer, the chain rule is used to calculate the gradient of the loss function for each parameter. Through back propagation, the gradient is transferred from the output layer to the input layer layer by layer to calculate the gradient value of each parameter. The gradient descent algorithm is used to update the parameters of the neural network according to the calculated parameter gradient. Repeat until a predetermined number of iterations or convergence condition of the loss function is reached; in each iteration, different training samples or sample batches are used to update the parameters to increase the randomness and generalization ability of training.

5. The intelligent evaluation method for power project investment based on artificial intelligence algorithm according to claim 3 is characterized in that: Rating models are validated and evaluated, including: The input data of the validation set is forward propagated through the trained rating model to obtain the output of the rating model. According to the output of the rating model, classification or prediction operations are performed to obtain the prediction results of the rating model for the validation set samples. Compare the prediction results of the rating model with the true labels of the validation set and calculate the accuracy of the model on the validation set; Based on the calculated accuracy, the performance of the rating model on the validation set is evaluated; the higher the accuracy, the better the performance of the rating model in predicting ratings; based on the evaluation results of the validation set, the rating model is adjusted and optimized.

6. The intelligent evaluation method for power project investment based on artificial intelligence algorithm according to claim 1 is characterized in that: Users can view the rating results and related indicators of power project companies through the interface or report.

7. The intelligent evaluation method for power project investment based on artificial intelligence algorithm according to claim 1 is characterized in that: The process of presenting the rating results to users in a visual form includes: Provide an arrangement area and condition configuration area for rating prediction results, and add elements related to rating prediction results through the arrangement area; present the rating prediction results to users in the form of a visual bar chart, which contains the rating results and related indicators of the power project enterprise; The visual rating results are embedded into the web page or application interface; users can input the power project enterprise information or select the power project enterprise of interest through the interactive operation of the interface to view its rating results and related indicators; Generate a report containing power project enterprise rating results and related indicators based on user needs.

8. The intelligent evaluation method for power project investment based on artificial intelligence algorithm according to claim 7 is characterized in that: By configuring the basic information, filtering conditions, aggregation conditions and sorting rules of the elements in the conditional configuration area, the visual configuration of the bar chart association analysis can be obtained.

9. The intelligent evaluation method for power project investment based on artificial intelligence algorithm according to claim 7 is characterized in that: The report includes a detailed description of the rating results, an explanation and analysis of the rating indicators, and a comparison with other power project companies or the industry average.

10. An intelligent evaluation system for power project investment based on artificial intelligence algorithm, which is applied to the intelligent evaluation method for power project investment based on artificial intelligence algorithm as claimed in any one of claims 1 to 9, characterized in that: The intelligent evaluation system for power project investment based on artificial intelligence algorithm includes: The power project investment intelligent post-evaluation subsystem is used to intelligently evaluate the investment effect of power projects through artificial intelligence algorithms, providing a comprehensive and accurate evaluation of investment effects; Feasibility study report review subsystem, used to review the feasibility study report of the power project; The wind turbine intelligent operation and maintenance evaluation subsystem is used to evaluate the intelligent operation and maintenance of the wind turbine. By analyzing and evaluating the operating data of the wind turbine, it provides information on the fan performance, operating status, and fault diagnosis.

Images