Substation construction project settlement review risk management method and system

Abstract

The present invention provides a method and system for risk control of settlement review of substation construction projects, which relates to the field of electric power technology. The specific scheme is: analyze the cost results data of completed power grid construction projects to obtain a key indicator system, summarize historical review issues to form a typical review case library, use positioning acquisition and image recognition technology to obtain key feature parameters of the site, and cluster to generate a typical engineering feature model. Comprehensively review the typical case library and the engineering feature model to provide a differentiated review strategy for power grid construction projects. This strategy enables the review model to review the cost data of the typical case library and the newly built power grid construction project by comparing them, and automatically generate a review report. In addition, since the review model includes a risk point prediction model, an accident tree analysis, and a risk prevention map, the review report can predict the potential risks in the implementation of the newly built power grid construction project, analyze the causes of the risks, and thus provide targeted suggestions.

Description

Technical Field

[0001] The present invention relates to the field of electric power technology, and in particular to a method, system, electronic equipment and storage medium for risk management and control of settlement review of a substation construction project. Background Art

[0002] There are many technical difficulties and contradictions in the settlement review of power grid projects.

[0003] The first is the construction of the settlement indicator system. It is necessary to adopt scientific methodology and data mining technology to extract key indicators from the massive general cost results, and to build specific indicator systems for different typical scenarios. Secondly, the sorting out of settlement review issues and case extraction, summarizing the rules from a large amount of historical data and cases, and providing a reference for subsequent risk prevention. Thirdly, in the on-site survey of the settlement review, it is necessary to use digital measurement technologies such as positioning acquisition and image recognition to improve the authenticity and accuracy of the review. Finally, due to the complexity and diversity of power grid projects, in the design and implementation of technical solutions, it is also necessary to consider the differentiated needs of different regions, different voltage levels, and different construction modes, coordinate the relationship between technological innovation and business applications, and ensure that the research results can truly serve the review practice and help the high-quality development of power grid projects. Summary of the invention

[0004] The present invention provides a substation construction project settlement review risk management method, system, electronic equipment and storage medium, which can solve at least one of the above problems.

[0005] The present invention provides a substation construction project settlement review risk management method, comprising:

[0006] Using data mining algorithms, we mine the historical cost results of power grid construction projects under various scenarios of substations to obtain multiple key settlement indicators.

[0007] Determining multiple valid settlement indicators based on indicators whose proportion of typical scenarios covered by the multiple key settlement indicators exceeds a preset threshold;

[0008] Performing problem statistics and classification on the historical settlement review data for reviewing the historical cost results data to obtain a settlement review problem case library;

[0009] By using image recognition technology, the on-site images of power grid construction projects in various scene types are analyzed to obtain multiple key feature parameters;

[0010] Clustering and classifying the multiple valid settlement indicators, the settlement review problem case library, and the multiple key characteristic parameters to obtain a typical characteristic model of the power grid construction project under each scenario type;

[0011] Determine differentiated review strategies and rule bases based on typical feature models of power grid construction projects under various scenario types;

[0012] Performing machine learning on the multiple valid settlement indicators, the settlement review problem case library, the multiple key feature parameters, and the differentiated review strategies and rule library to obtain an intelligent review model, wherein the intelligent review model includes a risk prevention knowledge graph generation model and a risk prediction model;

[0013] When the current cost result data of the power grid construction project is received, the current cost result data is reviewed based on the intelligent review model to obtain a review report on the cost result data, wherein the review report includes a description of problems, cause analysis and rectification suggestions, as well as risk prevention tips regarding the power grid construction project.

[0014] According to another aspect of the present invention, a substation construction project settlement review risk management and control system is provided, comprising:

[0015] The indicator mining module is used to mine the historical cost results data of power grid construction projects under various scenario types for substations using data mining algorithms to obtain multiple key settlement indicators;

[0016] An effective indicator determination module, used to determine a plurality of effective settlement indicators based on indicators whose proportion of typical scenarios covered by the plurality of key settlement indicators exceeds a preset threshold;

[0017] A case library determination module is used to statistically classify the problems of the historical settlement review data for reviewing the historical cost results data to obtain a settlement review problem case library;

[0018] A key feature determination module is used to analyze the on-site images of power grid construction projects under various scene types through image recognition technology to obtain multiple key feature parameters;

[0019] A typical model determination module is used to cluster and classify the multiple valid settlement indicators, the settlement review problem case library and the multiple key characteristic parameters to obtain a typical characteristic model of the power grid construction project under each scenario type;

[0020] A rule base determination module is used to determine the differentiated review strategy and rule base based on the typical characteristic models of power grid construction projects under various scenario types;

[0021] A review model determination module, used to perform machine learning on the multiple valid settlement indicators, the settlement review problem case library, the multiple key feature parameters, and the differentiated review strategies and rule library to obtain an intelligent review model, wherein the intelligent review model includes a risk prevention knowledge graph generation model and a risk prediction model;

[0022] The power grid project review module is used to review the current cost result data of the power grid construction project based on the intelligent review model upon receiving the current cost result data, and obtain a review report on the cost result data, wherein the review report includes a description of the problems, cause analysis and rectification suggestions, as well as risk prevention tips about the power grid construction project.

[0023] According to another aspect of the present invention, there is provided an electronic device, comprising:

[0024] at least one processor; and

[0025] a memory communicatively connected to the at least one processor; wherein,

[0026] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute any substation construction project settlement review risk management method in the embodiments of the present invention.

[0027] According to another aspect of the present invention, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute any substation construction project settlement review risk management method in the embodiments of the present invention.

[0028] By adopting the technical solution of the present invention, by analyzing the historical cost results data of power grid construction projects, multiple key settlement indicators are obtained, historical review issues are summarized to form a case library, and key feature parameters of the site are obtained by using positioning acquisition and image recognition technology. In combination with these data, a clustering algorithm is used to establish a typical engineering feature model. On this basis, the present invention designs a differentiated review strategy, so that in actual application, the cost results data of the power grid construction project is input into an intelligent review model based on deep learning for automated review, and a review report can be generated. In addition, since the review model includes risk prevention knowledge graph generation and risk prediction models, the potential risks of power grid construction projects can be predicted and prompted in the review report. The embodiment of the present invention realizes the intelligent and precise review of power grid project settlement, significantly improves the review efficiency and quality, and provides strong support for improving the cost management level of power grid projects.

[0029] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0031] Figure 1 It is a flow chart of a method for risk control of substation construction project settlement review according to an embodiment of the present invention;

[0032] Figure 2 It is a structural block diagram of a substation construction project settlement review risk management and control system according to an embodiment of the present invention;

[0033] Figure 3 is a block diagram of an electronic device for implementing an embodiment of the present invention. DETAILED DESCRIPTION

[0034] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0035] Figure 1 It is a flow chart of a method for risk control of substation construction project settlement review according to an embodiment of the present invention.

[0036] like Figure 1 As shown in the figure, the risk control methods for substation construction project settlement review can include:

[0037] S110, using a data mining algorithm to mine historical cost results data of power grid construction projects under various scenario types for substations, and obtain multiple key settlement indicators;

[0038] S120, determining multiple valid settlement indicators based on indicators whose proportion of typical scenarios covered by the multiple key settlement indicators exceeds a preset threshold;

[0039] S130, statistically classifying the problems of the historical settlement review data for reviewing the historical cost results data, and obtaining a settlement review problem case library;

[0040] S140, analyzing the on-site images of the power grid construction project under various scene types through image recognition technology to obtain multiple key feature parameters;

[0041] S150, clustering and classifying multiple valid settlement indicators, settlement review problem case libraries, and multiple key characteristic parameters to obtain typical characteristic models of power grid construction projects under various scenario types;

[0042] S160, based on the typical characteristic models of power grid construction projects under various scenario types, determine the differentiated review strategy and rule base;

[0043] S170, performing machine learning on multiple valid settlement indicators, a settlement review problem case library, multiple key feature parameters, and a differentiated review strategy and rule library to obtain an intelligent review model, wherein the intelligent review module includes a risk prevention knowledge graph generation model and a risk prediction model;

[0044] S180, when receiving the current cost result data of the power grid construction project, the current cost result data is reviewed based on the intelligent review model to obtain a review report on the cost result data, wherein the review report includes a description of the problems, cause analysis and rectification suggestions, as well as risk prevention tips about the power grid construction project.

[0045] Exemplarily, the data mining algorithm may include an Apriori association rule mining algorithm, a CART decision tree algorithm, and the like.

[0046] It can be understood that the historical cost results data refers to the cost data of the completed power grid construction projects, and the current cost results data refers to the cost data of the newly completed power grid construction projects.

[0047] For example, the key settlement indicators may be material costs and labor costs of the high-voltage line project.

[0048] For example, power grid construction projects under different scenario types may have different geographical areas, different voltage levels, or different construction modes.

[0049] Exemplarily, the typical scene may be a typical scene corresponding to each of the above-mentioned scene categories. The typical scene may be understood as a power grid construction project of a certain scene category.

[0050] Exemplarily, the risk prevention knowledge graph generation model is used to display the correlation between the risk settlement issues of the power grid construction project, and the risk prediction model is used to predict the possible risk settlement issues. For example, the risk prediction model may include a risk point prediction model and an accident tree analysis model.

[0051] According to the above implementation mode, by analyzing the historical cost results data of the power grid construction project, multiple key settlement indicators are obtained, historical review issues are summarized to form a case library, and the key feature parameters of the site are obtained by using positioning acquisition and image recognition technology. In combination with these data, a clustering algorithm is used to establish a typical engineering feature model. On this basis, the present invention designs a differentiated review strategy, so that in actual application, the cost results data of the power grid construction project is input into the intelligent review model based on deep learning for automatic review, and a review report can be generated. In addition, since the review model includes risk prevention knowledge graph generation and risk prediction model, the potential risks of the power grid construction project can be predicted and prompted in the review report. The embodiment of the present invention realizes the intelligent and precise review of the settlement of power grid projects, significantly improves the review efficiency and quality, and provides strong support for improving the cost management level of power grid projects.

[0052] In one embodiment, a data mining algorithm is used to mine historical cost results data of power grid construction projects under various scenario types for substations to obtain multiple key settlement indicators, including: using an Apriori association rule mining algorithm to mine historical cost results data to obtain association rules between multiple settlement indicators; constructing a settlement indicator decision tree based on the settlement indicators involved in the association rules; searching in the settlement indicator decision tree based on scenario features corresponding to power grid construction projects under various scenario types to obtain multiple key settlement indicators.

[0053] For example, the massive historical cost results data of power grid construction projects are first preprocessed. For example, data cleaning techniques such as the outlier detection algorithm LOF (Local Outlier Factor) are used to identify and remove noise data and outliers to improve data quality. Then, the cleaned data is standardized, and the Z-score standardization method is used to scale the data to a standard normal distribution with a mean of 0 and a variance of 1, eliminating the dimensional effects between different indicators.

[0054] For example, the Apriori association rule mining algorithm is used, with the minimum support set to 05 and the minimum confidence set to 8, to mine the association rules between settlement indicators in typical scenarios such as different regions, voltage levels, and construction modes. For example, the correlation between "material costs" and "labor costs" in the "high-voltage line engineering" scenario reaches 9, and an association rule can be obtained: "material costs" and "labor costs" are associated in the "high-voltage line engineering" scenario.

[0055] Exemplarily, the CART decision tree algorithm is used to learn the settlement indicators involved in the association rules and construct a settlement indicator decision tree.

[0056] Understandably, multiple key settlement indicators can be searched for each scenario type, and a settlement indicator system can be constructed based on multiple key settlement indicators corresponding to multiple scenario types.

[0057] According to the above implementation mode, a data mining algorithm is used to mine the historical cost results data of power grid construction projects under various scenario types for substations to obtain multiple key settlement indicators, thereby constructing a settlement indicator system for power grid construction projects.

[0058] In one embodiment, based on the indicators in which the proportion of typical scenarios covered by multiple key settlement indicators exceeds a preset threshold, multiple effective settlement indicators are determined, including: when the ratio between the number of scenario types corresponding to the power grid construction project covered by the key settlement indicators and the total number of all scenario types exceeds a preset threshold, the key settlement indicator is used as a valid settlement indicator.

[0059] Exemplarily, when the ratio between the number of scenario types corresponding to the power grid construction project covered by the key settlement indicator and the total number of all scenario types does not exceed a preset threshold, it is considered to be an invalid settlement indicator.

[0060] In this way, multiple valid settlement indicators and multiple invalid settlement indicators can be obtained.

[0061] Exemplarily, the proportion of these two indicators is counted. If the proportion of the effective settlement indicator is less than the preset proportion threshold, return to execute the above step of "using data mining algorithm to mine the historical cost results data of power grid construction projects under various scenario types for substations to obtain multiple key settlement indicators" until the proportion of the effective settlement indicator is greater than the preset proportion threshold.

[0062] For example, when returning to execute the step of "using a data mining algorithm to mine the historical cost results data of power grid construction projects under various scenario types for substations to obtain multiple key settlement indicators", the support and confidence parameters of the Apriori association rule mining algorithm and the tree depth and other parameters of the CART decision tree algorithm can be adjusted.

[0063] Exemplarily, based on the multiple effective settlement indicators obtained after optimization, a settlement indicator system for power grid construction projects is constructed.

[0064] For example, the optimized settlement indicator system is applied to actual power grid project cost settlement cases. The effect of improving settlement efficiency and accuracy is evaluated through AB testing and other methods, and the settlement indicator system is further optimized for scenarios with unsatisfactory results. Finally, the settlement indicator system is stored in NoSQL databases such as MongoDB in JSON format to form a standardized settlement indicator model, and a regular update mechanism is established to automatically update the indicator model every quarter based on the newly added cost data, and use incremental learning algorithms such as Hoeffding Tree to dynamically optimize the decision tree to maintain the timeliness and accuracy of the indicator system.

[0065] According to the above implementation, based on the indicators whose proportion of typical scenarios covered by multiple key settlement indicators exceeds a preset threshold, multiple effective settlement indicators are determined, so that a settlement indicator system for power grid construction projects can be constructed based on multiple effective settlement indicators, thereby improving the accuracy of the settlement indicator system.

[0066] In one embodiment, historical settlement review data for reviewing historical cost results data are statistically classified to obtain a settlement review problem case library, including: using the FP-Growth algorithm to perform frequent pattern mining on the historical settlement review data to obtain multiple problem frequent occurrence patterns and multiple problem occurrence regular characteristics; clustering the multiple problem frequent occurrence patterns and multiple problem occurrence regular characteristics to obtain multiple first clustering clusters, wherein each first clustering cluster corresponds to a settlement problem; for each first clustering cluster, based on the problem frequent occurrence pattern and problem occurrence regular characteristics in the first clustering cluster, typical cases are extracted from the historical settlement review data to obtain typical cases of settlement problems corresponding to the first clustering cluster; based on the typical cases of settlement problems corresponding to each first clustering cluster, a settlement review problem case library is determined.

[0067] For example, first, we obtain the historical settlement review data of the past three years from the historical settlement review database, totaling 1.2 million records. We use data cleaning tools to pre-process the acquired data, remove invalid data with missing fields exceeding 5%, and unify the time format to "yyyy-MM-dd HH:mm:ss", and the amount format to retain two decimal places, and finally obtain a standardized settlement review data set.

[0068] Exemplarily, the Apriori association rule mining algorithm is used, with the minimum support set to 05 and the minimum confidence set to 8, to mine frequent patterns of problem occurrences, such as "medical institution A, medical project B→problem C (support 07, confidence 85)", and a total of 120 problem regularity features are obtained.

[0069] Exemplarily, the K-means clustering algorithm is used to automatically classify the problems, and the number of clusters k is set to 10, the number of iterations is 100, and the convergence threshold is 0.01, and 10 problem categories are obtained. For each problem category, typical cases are randomly selected from the settlement review data set at a ratio of 20%, forming a settlement review problem case library containing 12,000 cases.

[0070] According to the above implementation mode, the historical settlement review data for reviewing the historical cost results data can be statistically classified to obtain a settlement review problem case library.

[0071] In one embodiment, through image recognition technology, on-site images of power grid construction projects under various scene types are analyzed to obtain multiple key feature parameters, including: using a pre-built image segmentation model to segment each on-site image to obtain multiple key area images; using a pre-built image recognition model to perform feature recognition on each key area image to obtain multiple candidate key features; for each candidate key feature, using a preset classification model to identify the candidate key feature and obtain a recognition result corresponding to the candidate key feature; based on each candidate key feature and the corresponding recognition result, determining the recognition accuracy of the image recognition model; when the recognition accuracy is greater than a preset accuracy threshold, determining multiple key feature parameters based on multiple candidate key features.

[0072] For example, in the construction site management, the GPS positioning technology is first used to obtain the specific location coordinates of the site, such as (30°5′22′′N, 11°24′37′′E). Then, the site image corresponding to the positioning position is obtained.

[0073] For example, a convolutional neural network (CNN) model, such as the U-Net architecture, is used to segment the on-site images. For example, the U-Net model extracts image features through multi-layer convolution and pooling operations, and finally outputs a segmentation map to accurately mark key areas, such as construction areas, equipment storage areas, etc.

[0074] For example, the segmented image data is input into a pre-trained deep learning model, such as ResNet-50, to obtain multiple candidate key features. The model can be fine-tuned using a transfer learning method. By training on a dataset of engineering site images, the model parameters are adjusted to identify key feature parameters such as engineering scale (e.g., small, medium, large), equipment type (e.g., excavator, crane), and construction difficulty (e.g., low, medium, high).

[0075] Exemplarily, each candidate key feature is classified by a support vector machine (SVM) model to obtain a recognition result corresponding to each candidate key feature, wherein the SVM may use a Gaussian kernel function, and set C=0 and gamma=1.

[0076] For example, if the classification accuracy reaches 98%, since it exceeds the preset threshold of 95%, multiple candidate key features can be used as the above multiple key feature parameters. If the classification accuracy does not reach the preset threshold of 95%, an incremental learning method can be used to optimize the image segmentation model and image recognition model using 1,000 newly collected engineering site images, adjust the number of convolutional layer filters and activation functions, and stop training until the recognition accuracy is increased to 95% after training.

[0077] According to the above implementation, multiple key feature parameters can be obtained by analyzing the on-site images of power grid construction projects in various scene types through image recognition technology.

[0078] In one embodiment, multiple valid settlement indicators, settlement review problem case libraries, and multiple key characteristic parameters are clustered and classified to obtain typical characteristic models of power grid construction projects under various scenario types, including: clustering multiple valid settlement indicators, settlement review problem case libraries, and multiple key characteristic parameters to obtain multiple second clustering clusters; using a scenario type classification model to classify and identify each second clustering cluster to obtain a scenario type corresponding to each second clustering cluster; based on the second clustering clusters corresponding to each scenario type, respectively determine the typical characteristic model of the power grid construction project under each scenario type.

[0079] Exemplarily, each second cluster may include at least one effective settlement indicator, at least one settlement review problem case, and at least one key feature parameter.

[0080] Exemplarily, the scene type classification model is a pre-trained model, which can be trained and generated using a random forest or a decision tree.

[0081] Exemplarily, machine learning may be performed on the second cluster corresponding to each scene type to obtain a corresponding typical feature model.

[0082] According to the above implementation mode, multiple valid settlement indicators, settlement review problem case libraries and multiple key characteristic parameters are clustered and classified to obtain a typical characteristic model of the power grid construction project under each scenario type.

[0083] In one embodiment, based on typical feature models of power grid construction projects under various scenario types, differentiated review strategies and rule bases are determined, including: based on typical feature models of power grid construction projects under various scenario types, review strategies and rules corresponding to each scenario type are determined, wherein the review strategies and rules include at least one risk threshold condition corresponding to a rule; based on the review strategies and rules corresponding to each scenario type, differentiated review strategies and rule bases are determined.

[0084] Exemplarily, the review strategy and rules may include that if the value of a certain settlement indicator exceeds a preset risk threshold condition, then it is determined that the project corresponding to the settlement indicator has a risk settlement problem. Alternatively, the review strategy and rules may also include that if the similarity between the project characteristics of the power grid construction project and the characteristics of the typical cases in the typical case library exceeds a preset similarity threshold, then it is considered that the power grid construction project has a risk settlement problem.

[0085] Exemplarily, the differentiated review policies and rule base may include review policies and rules divided by scenario type.

[0086] According to the above implementation, differentiated review strategies and rule bases are determined based on typical feature models of power grid construction projects under various scenario types.

[0087] Figure 2 It is a structural block diagram of a substation construction project settlement review risk management and control system according to an embodiment of the present invention.

[0088] like Figure 2 As shown, the substation construction project settlement review risk management and control system may include:

[0089] The indicator mining module 210 is used to mine the historical cost results data of the power grid construction project under various scenario types for the substation using a data mining algorithm to obtain multiple key settlement indicators;

[0090] An effective indicator determination module 220, configured to determine a plurality of effective settlement indicators based on indicators whose proportion of typical scenarios covered by the plurality of key settlement indicators exceeds a preset threshold;

[0091] A case library determination module 230 is used to statistically classify the historical settlement review data for reviewing the historical cost achievement data to obtain a settlement review problem case library;

[0092] The key feature determination module 240 is used to analyze the on-site images of the power grid construction project under various scene types by using image recognition technology to obtain multiple key feature parameters;

[0093] A typical model determination module 250 is used to cluster and classify the multiple valid settlement indicators, the settlement review problem case library and the multiple key characteristic parameters to obtain a typical characteristic model of the power grid construction project under each scenario type;

[0094] A rule base determination module 260 is used to determine a differentiated review strategy and a rule base based on a typical feature model of a power grid construction project under each scenario type;

[0095] The review model determination module 270 is used to perform machine learning on the multiple valid settlement indicators, the settlement review problem case library, the multiple key feature parameters, and the differentiated review strategy and rule library to obtain an intelligent review model, wherein the intelligent review model includes a risk prevention knowledge graph generation model and a risk prediction model;

[0096] The power grid project review module 280 is used to review the current cost result data of the power grid construction project based on the intelligent review model upon receiving the current cost result data, and obtain a review report on the cost result data, wherein the review report includes a description of the problems, cause analysis and rectification suggestions, as well as risk prevention tips regarding the power grid construction project.

[0097] In one embodiment, the indicator mining module includes:

[0098] An association rule determination unit, used to mine the historical cost results data using an Apriori association rule mining algorithm to obtain association rules between multiple settlement indicators;

[0099] A decision tree determination unit, used to construct a settlement indicator decision tree based on the settlement indicators involved in the association rules;

[0100] The key indicator determination unit is used to search in the settlement indicator decision tree based on the scenario characteristics corresponding to the power grid construction project under each scenario type to obtain multiple key settlement indicators.

[0101] In one implementation, the effective index determination module is specifically used to:

[0102] When the ratio between the number of scenario types corresponding to the power grid construction project covered by the key settlement indicator and the total number of all scenario types exceeds a preset threshold, the key settlement indicator is used as the effective settlement indicator.

[0103] In one embodiment, the case library determination module includes:

[0104] A pattern mining unit, used to perform frequent pattern mining on the historical settlement review data using an FP-Growth algorithm to obtain a plurality of frequent occurrence patterns of problems and a plurality of regular occurrence characteristics of problems;

[0105] A first clustering unit is used to cluster the multiple problem frequent occurrence patterns and the multiple problem occurrence regularity characteristics to obtain multiple first clusters, wherein each of the first clusters corresponds to a settlement problem;

[0106] A case extraction unit, configured to extract typical cases from the historical settlement review data for each of the first clusters based on the frequent occurrence patterns and regular characteristics of the problems in the first clusters, to obtain typical cases of settlement problems corresponding to the first clusters;

[0107] The case library determining unit is used to determine the settlement review problem case library based on typical cases of settlement problems corresponding to each of the first clusters.

[0108] In one embodiment, the key feature determination module includes:

[0109] An image segmentation unit, used to use a pre-built image segmentation model to perform segmentation processing on each of the scene images to obtain a plurality of key area images;

[0110] A special recognition unit is used to use a pre-built image recognition model to perform feature recognition on each of the key area images to obtain a plurality of candidate key features;

[0111] A feature classification unit, used for identifying each of the candidate key features by using a preset classification model, and obtaining an identification result corresponding to the candidate key feature;

[0112] An accuracy determination unit, used to determine the recognition accuracy of the image recognition model based on each of the candidate key features and the corresponding recognition results;

[0113] The key feature determination unit is used to determine the multiple key feature parameters based on the multiple candidate key features when the recognition accuracy is greater than a preset accuracy threshold.

[0114] In one implementation, the typical model determination module includes:

[0115] A second clustering unit is used to cluster the multiple valid settlement indicators, the settlement review problem case library and the multiple key feature parameters to obtain multiple second clusters;

[0116] a type identification unit, configured to classify and identify each of the second clusters using a scene type classification model to obtain a scene type corresponding to each of the second clusters;

[0117] The model determination unit is used to determine the typical characteristic model of the power grid construction project under each scenario type based on the second clustering cluster corresponding to each scenario type.

[0118] In one embodiment, the rule base determination module includes:

[0119] A rule determination unit, configured to determine the review strategy and rules corresponding to each scenario type based on a typical feature model of a power grid construction project under each scenario type, wherein the review strategy and rules include a risk threshold condition corresponding to at least one rule;

[0120] The rule base determination unit is used to determine the differentiated review strategy and rule base based on the review strategy and rules corresponding to each scenario type.

[0121] For the description of specific functions and examples of each module and submodule of the system in the embodiment of the present invention, reference can be made to the relevant description of the corresponding steps in the above method embodiment, which will not be repeated here.

[0122] According to an embodiment of the present invention, the above method of the present invention can be applied to an electronic device and a readable storage medium.

[0123] Figure 3 A schematic block diagram of an example electronic device 800 that can be used to implement an embodiment of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0124] like Figure 3 As shown, the device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 to a random access memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the device 800 can also be stored. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0125] A number of components in the device 800 are connected to the I / O interface 805, including: an input unit 806, such as a keyboard, a mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a disk, an optical disk, etc.; and a communication unit 809, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 809 allows the device 800 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0126] The computing unit 801 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 801 performs the various methods and processes described above, such as a substation construction project settlement review risk control method. For example, in some embodiments, the substation construction project settlement review risk control method may be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as a storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 800 via the ROM 802 and / or the communication unit 809. When the computer program is loaded into the RAM 803 and executed by the computing unit 801, one or more steps of the substation construction project settlement review risk control method described above may be performed. Alternatively, in other embodiments, the computing unit 801 may be configured to execute the substation construction project settlement review risk management method in any other appropriate manner (eg, by means of firmware).

[0127] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0128] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0129] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or device, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0130] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0131] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0132] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0133] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0134] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the principles of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A method for risk control of settlement review of substation construction project, characterized in that: include: Using data mining algorithms, we mine the historical cost results of power grid construction projects under various scenarios of substations to obtain multiple key settlement indicators. Determining multiple valid settlement indicators based on indicators whose proportion of typical scenarios covered by the multiple key settlement indicators exceeds a preset threshold; Performing problem statistics and classification on the historical settlement review data for reviewing the historical cost results data to obtain a settlement review problem case library, including: using the FP-Growth algorithm to perform frequent pattern mining on the historical settlement review data to obtain multiple problem frequent occurrence patterns and multiple problem occurrence regularity characteristics; clustering multiple problem frequent occurrence patterns and multiple problem occurrence regularity characteristics to obtain multiple first clustering clusters, wherein each of the first clustering clusters corresponds to a settlement problem; for each of the first clustering clusters, extracting typical cases from the historical settlement review data based on the problem frequent occurrence patterns and problem occurrence regularity characteristics in the first clustering clusters to obtain typical cases of settlement problems corresponding to the first clustering clusters; determining the settlement review problem case library based on the typical cases of settlement problems corresponding to each of the first clustering clusters; By using image recognition technology, we analyze the on-site images of power grid construction projects in various scene types and obtain multiple key feature parameters; Clustering and classifying the multiple valid settlement indicators, the settlement review problem case library, and the multiple key characteristic parameters to obtain typical characteristic models of power grid construction projects under various scenario types, including: clustering the multiple valid settlement indicators, the settlement review problem case library, and the multiple key characteristic parameters to obtain multiple second clusters; using a scenario type classification model to classify and identify each of the second clusters to obtain a scenario type corresponding to each of the second clusters; based on the second clusters corresponding to each scenario type, respectively determining the typical characteristic models of power grid construction projects under each scenario type; Determine differentiated review strategies and rule bases based on typical feature models of power grid construction projects under various scenario types; Performing machine learning on the multiple valid settlement indicators, the settlement review problem case library, the multiple key feature parameters, and the differentiated review strategies and rule library to obtain an intelligent review model, wherein the intelligent review model includes a risk prevention knowledge graph generation model and a risk prediction model; When the current cost result data of the power grid construction project is received, the current cost result data is reviewed based on the intelligent review model to obtain a review report on the cost result data, wherein the review report includes a description of problems, cause analysis and rectification suggestions, as well as risk prevention tips regarding the power grid construction project.

2. The method according to claim 1, characterized in that The data mining algorithm is used to mine the historical cost results data of power grid construction projects under various scenario types for substations to obtain multiple key settlement indicators, including: Using the Apriori association rule mining algorithm, the historical cost results data is mined to obtain association rules between multiple settlement indicators; Based on the settlement indicators involved in the association rules, construct a settlement indicator decision tree; Based on the scenario characteristics corresponding to the power grid construction project under each of the scenario types, a search is performed in the settlement indicator decision tree to obtain a plurality of the key settlement indicators.

3. The method according to claim 1, characterized in that The determining of multiple valid settlement indicators based on the indicator whose proportion of typical scenarios covered by the multiple key settlement indicators exceeds a preset threshold includes: When the ratio between the number of scenario types corresponding to the power grid construction project covered by the key settlement indicator and the total number of all scenario types exceeds a preset threshold, the key settlement indicator is used as the effective settlement indicator.

4. The method according to claim 1, characterized in that: The image recognition technology is used to analyze the on-site images of power grid construction projects in various scene types to obtain multiple key feature parameters, including: Using a pre-built image segmentation model, segment each of the scene images to obtain a plurality of key area images; Using a pre-built image recognition model, perform feature recognition on each of the key area images to obtain a plurality of candidate key features; For each of the candidate key features, a preset classification model is used to identify the candidate key feature to obtain an identification result corresponding to the candidate key feature; Determining the recognition accuracy of the image recognition model based on each of the candidate key features and the corresponding recognition results; When the recognition accuracy is greater than a preset accuracy threshold, the multiple key feature parameters are determined based on the multiple candidate key features.

5. The method according to claim 1, characterized in that The typical characteristic models of power grid construction projects under various scenario types are used to determine differentiated review strategies and rule bases, including: Based on the typical characteristic models of power grid construction projects under various scenario types, determine the review strategies and rules corresponding to each scenario type, wherein the review strategies and rules include at least one risk threshold condition corresponding to a rule; Based on the review strategies and rules corresponding to each scenario type, the differentiated review strategy and rule base are determined.

6. A substation construction project settlement review risk management and control system, characterized in that: include: The indicator mining module is used to mine the historical cost results data of power grid construction projects under various scenario types for substations using data mining algorithms to obtain multiple key settlement indicators; An effective indicator determination module, used to determine a plurality of effective settlement indicators based on indicators whose proportion of typical scenarios covered by the plurality of key settlement indicators exceeds a preset threshold; A case library determination module is used to statistically classify the problems of the historical settlement review data for reviewing the historical cost results data to obtain a settlement review problem case library; A key feature determination module is used to analyze the on-site images of power grid construction projects under various scene types through image recognition technology to obtain multiple key feature parameters; A typical model determination module is used to cluster and classify the multiple valid settlement indicators, the settlement review problem case library and the multiple key characteristic parameters to obtain a typical characteristic model of the power grid construction project under each scenario type; A rule base determination module is used to determine the differentiated review strategy and rule base based on the typical characteristic models of power grid construction projects under various scenario types; A review model determination module, used to perform machine learning on the multiple valid settlement indicators, the settlement review problem case library, the multiple key feature parameters, and the differentiated review strategies and rule library to obtain an intelligent review model, wherein the intelligent review model includes a risk prevention knowledge graph generation model and a risk prediction model; A power grid project review module, for reviewing the current cost result data of the power grid construction project based on the intelligent review model upon receiving the current cost result data, and obtaining a review report on the cost result data, wherein the review report includes a description of the problem, cause analysis and rectification suggestions, and risk prevention tips for the power grid construction project; Wherein, the case library determination module includes: A pattern mining unit, used to perform frequent pattern mining on the historical settlement review data using an FP-Growth algorithm to obtain a plurality of frequent occurrence patterns of problems and a plurality of regular occurrence characteristics of problems; A first clustering unit is used to cluster the multiple problem frequent occurrence patterns and the multiple problem occurrence regularity characteristics to obtain multiple first clusters, wherein each of the first clusters corresponds to a settlement problem; A case extraction unit, configured to extract typical cases from the historical settlement review data for each of the first clusters based on the frequent occurrence patterns and regular characteristics of the problems in the first clusters, to obtain typical cases of settlement problems corresponding to the first clusters; A case library determining unit, configured to determine the settlement review problem case library based on typical cases of settlement problems corresponding to each of the first clusters; Wherein, the typical model determination module includes: A second clustering unit is used to cluster the multiple valid settlement indicators, the settlement review problem case library and the multiple key feature parameters to obtain multiple second clusters; a type identification unit, configured to classify and identify each of the second clusters using a scene type classification model to obtain a scene type corresponding to each of the second clusters; The model determination unit is used to determine the typical characteristic model of the power grid construction project under each scenario type based on the second clustering cluster corresponding to each scenario type.

7. The system according to claim 6, characterized in that The indicator mining module includes: An association rule determination unit, used to mine the historical cost results data using an Apriori association rule mining algorithm to obtain association rules between multiple settlement indicators; A decision tree determination unit, used to construct a settlement indicator decision tree based on the settlement indicators involved in the association rules; The key indicator determination unit is used to search in the settlement indicator decision tree based on the scenario characteristics corresponding to the power grid construction project under each scenario type to obtain multiple key settlement indicators.

8. The system according to claim 6, characterized in that The effective index determination module is specifically used for: When the ratio between the number of scenario types corresponding to the power grid construction project covered by the key settlement indicator and the total number of all scenario types exceeds a preset threshold, the key settlement indicator is used as the effective settlement indicator.

9. The system according to claim 6, characterized in that The key feature determination module comprises: An image segmentation unit, used to use a pre-built image segmentation model to perform segmentation processing on each of the scene images to obtain a plurality of key area images; A special recognition unit is used to use a pre-built image recognition model to perform feature recognition on each of the key area images to obtain a plurality of candidate key features; A feature classification unit, used for identifying each of the candidate key features by using a preset classification model, and obtaining an identification result corresponding to the candidate key feature; An accuracy determination unit, used to determine the recognition accuracy of the image recognition model based on each of the candidate key features and the corresponding recognition results; The key feature determination unit is used to determine the multiple key feature parameters based on the multiple candidate key features when the recognition accuracy is greater than a preset accuracy threshold.

10. An electronic device comprising: at least one processor; as well as a memory communicatively coupled to the at least one processor; The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method described in any one of claims 1 to 5.

11. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-5.

Images