A power construction document verification method and device, an electronic device, and a storage medium

Abstract

Embodiments of the present application disclose a power construction document checking method and device, electronic equipment and a storage medium. The power construction document checking method comprises: obtaining a to-be-checked power construction project document, and determining at least one first to-be-checked document and at least one second to-be-checked document in the to-be-checked power construction project document; sequentially inputting each first to-be-checked document into a pre-trained document classification model to determine the document type of each first to-be-checked document through the document classification model; performing complete specification checking on each first to-be-checked document according to the document type of each first to-be-checked document; and performing document logic checking on each second to-be-checked document in the case that the complete specification checking result of each first to-be-checked document meets the complete specification checking condition. The technical solution of the embodiments of the present application can automatically check the power construction document, and improve the checking speed and accuracy of the power construction document.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of computer technology, and in particular to a method, apparatus, electronic device and storage medium for verifying power construction documents. Background Technology

[0002] Currently, every power construction project generates a large number of power construction documents during the construction process, including various document types. Each document type has different document formats and content. To facilitate project management and review of power construction projects, it is necessary to verify these power construction documents.

[0003] However, existing methods for verifying power construction documents are usually done manually, but manual verification is slow and prone to errors, resulting in low accuracy. Summary of the Invention

[0004] This invention provides a method, apparatus, electronic device, and storage medium for verifying power construction documents, which can automatically verify power construction documents and improve the verification speed and accuracy.

[0005] According to one aspect of the present invention, a method for verifying power construction documents is provided, comprising:

[0006] Obtain the power construction project documents to be verified, and determine at least one first document to be verified and at least one second document to be verified from the power construction project documents to be verified; wherein, the second document to be verified is a document with a set logical relationship among the first documents to be verified;

[0007] Each of the first documents to be verified is sequentially input into a pre-trained document classification model to determine the document type of each of the first documents to be verified through the document classification model.

[0008] Based on the document type of each of the first documents to be verified, a complete and standardized verification is performed on each of the first documents to be verified.

[0009] If the complete specification verification results of each of the first documents to be verified meet the complete specification verification conditions, then document logic verification is performed on each of the second documents to be verified.

[0010] According to another aspect of the present invention, a power construction document verification device is provided, comprising:

[0011] The document to be verified module is used to acquire power construction project documents to be verified, and to determine at least one first document to be verified and at least one second document to be verified from the power construction project documents to be verified; wherein, the second document to be verified is a document with a set logical relationship among the first documents to be verified.

[0012] The document type determination module is used to sequentially input each of the first documents to be verified into a pre-trained document classification model, so as to determine the document type of each of the first documents to be verified through the document classification model;

[0013] The complete specification verification module is used to perform complete specification verification on each of the first documents to be verified according to the document type of each document to be verified.

[0014] The document logic verification module is used to perform document logic verification on each of the second documents to be verified, provided that the complete specification verification results of each of the first documents to be verified meet the complete specification verification conditions.

[0015] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0016] At least one processor; and

[0017] A memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the power construction document verification method according to any embodiment of the present invention.

[0019] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions, the computer instructions being configured to cause a processor to execute and implement the power construction document verification method according to any embodiment of the present invention.

[0020] The technical solution of this invention obtains the power construction project documents to be verified, determines the first and second documents to be verified, and sequentially inputs each first document to be verified into a pre-trained document classification model to determine the document type of each first document. Based on the document type, a complete specification verification is performed on each first document to be verified. Then, if the complete specification verification results of each first document meet the complete specification verification conditions, a document logic verification is performed on each second document to be verified. This solves the problems of slow speed and low accuracy caused by manual verification of existing power construction documents, enabling automatic verification of power construction documents and improving the verification speed and accuracy.

[0021] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is a flowchart of a power construction document verification method provided in Embodiment 1 of the present invention;

[0024] Figure 2 This is a flowchart of a power construction document verification method provided in Embodiment 2 of the present invention;

[0025] Figure 3 This is an example architecture diagram of a power construction document verification device provided in Embodiment 2 of the present invention;

[0026] Figure 4 This is a schematic diagram of a power construction document verification device provided in Embodiment 3 of the present invention;

[0027] Figure 5 This is a schematic diagram of the structure of an electronic device that implements the power construction document verification method of this invention. Detailed Implementation

[0028] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0029] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0030] Example 1

[0031] Figure 1 This is a flowchart of a power construction document verification method provided in Embodiment 1 of the present invention. This embodiment is applicable to improving the verification speed and accuracy of power construction documents. The method can be executed by a power construction document verification device, which can be implemented through software and / or hardware, and is generally directly integrated into the electronic device executing this method. This electronic device can be a terminal device or a server device. The present invention does not limit the type of electronic device executing the power construction document verification method. Specifically, as shown... Figure 1 As shown, the power construction document verification method may specifically include the following steps:

[0032] S110. Obtain the power construction project documents to be verified, and determine at least one first document to be verified and at least one second document to be verified from the power construction project documents to be verified; wherein, the second document to be verified is a document in the first document to be verified that has a set logical relationship.

[0033] The power construction project documents to be verified can be any documents generated from the power construction project. It is understood that after the completion of a power construction project, multiple documents of different types may be generated, and all generated documents need to be verified. The first document to be verified can be any document within the power construction project documents that needs to be verified. The second document to be verified can be a document within the first document to be verified that has a defined logical relationship. This defined logical relationship can be a logical relationship set according to the actual needs of the power construction project, such as a time-based logical relationship or a content-based logical relationship, etc., and this embodiment of the invention does not impose any limitations on it.

[0034] In this embodiment of the invention, a power construction project document to be verified is obtained, and a first document and a second document to be verified are identified within the power construction project document. Specifically, the second document to be verified may be a document within the first document to be verified that has a defined logical relationship. It is understood that the number of the first document to be verified may be less than or equal to the total number of documents in the power construction project document to be verified. The number of the second document to be verified may be less than or equal to the number of the first document to be verified.

[0035] S120. Each of the first documents to be verified is sequentially input into a pre-trained document classification model to determine the document type of each of the first documents to be verified through the document classification model.

[0036] The document classification model can be a model that classifies the first document to be verified.

[0037] In this embodiment of the invention, after obtaining the power construction project documents to be verified and identifying at least one first document to be verified among them, each first document to be verified can be further input sequentially into a pre-trained document classification model to determine the document type of each first document to be verified. For example, document types may include supervision documents, planning documents, evaluation documents, construction documents, and acceptance documents, etc.

[0038] S130. Perform a complete specification verification on each of the first documents to be verified according to the document type of each first document to be verified.

[0039] Among them, complete specification verification can verify the completeness and standardization of the document.

[0040] In this embodiment of the invention, after determining the document type of each first document to be verified through a document classification model, a completeness and standardization verification can be further performed on each first document to be verified based on its document type. It is understood that documents of the same document type have the same requirements for completeness and standardization.

[0041] S140. If the complete specification verification results of each of the first documents to be verified meet the complete specification verification conditions, perform document logic verification on each of the second documents to be verified.

[0042] The completeness and standardization verification result can be the result obtained after verifying the completeness and standardization of the document. The completeness and standardization verification conditions can be the conditions for achieving completeness and standardization. That is, meeting the completeness and standardization verification conditions indicates that completeness and standardization are met. For example, the completeness and standardization verification result can be a numerical value; when this value reaches the threshold set by the completeness and standardization verification conditions, it can be determined that the document meets both completeness and standardization requirements.

[0043] In this embodiment of the invention, after performing a complete specification check on each first document to be checked according to its document type, it can be further determined whether the complete specification check result of each first document to be checked meets the complete specification check condition, and when the complete specification check result of each first document to be checked meets the complete specification check condition, a document logic check is performed on each second document to be checked.

[0044] The technical solution of this embodiment obtains the power construction project documents to be verified, identifies the first and second documents to be verified, and sequentially inputs each first document to be verified into a pre-trained document classification model. The document classification model determines the document type of each first document to be verified, and performs a complete specification verification on each first document to be verified based on its document type. Then, if the complete specification verification results of each first document to be verified meet the complete specification verification conditions, a document logic verification is performed on each second document to be verified. This solves the problems of slow speed and low accuracy caused by manual verification of existing power construction documents, and can automatically verify power construction documents, improving the verification speed and accuracy.

[0045] Example 2

[0046] Figure 2This is a flowchart of a power construction document verification method provided in Embodiment 2 of the present invention. This embodiment further refines the above-mentioned technical solutions, providing various specific optional implementation methods for sequentially inputting each of the first documents to be verified into a pre-trained document classification model to determine the document type of each of the first documents to be verified through the document classification model, performing a complete specification verification on each of the first documents to be verified based on the document type, and performing document logic verification on each of the second documents to be verified when the complete specification verification results of each of the first documents to be verified meet the complete specification verification conditions. The technical solution in this embodiment can be combined with various optional solutions in one or more of the above embodiments. Figure 2 As shown, the method may include the following steps:

[0047] S210. Obtain the power construction project documents to be verified, and determine at least one first document to be verified and at least one second document to be verified from the power construction project documents to be verified; wherein, the second document to be verified is a document in the first document to be verified that has a set logical relationship.

[0048] Optionally, before obtaining the power construction project documents to be verified, the process may further include: obtaining a first historical power construction document and performing single-character recognition on the first historical power construction document to obtain an identifiable document; determining the document type tag of the identifiable document; obtaining the document title in the identifiable document; obtaining the document content corresponding to the document title and performing semantic recognition on the document content to determine the content type of the document content; and constructing a standard document based on the document type tag, document title, and content type of the document content.

[0049] The first historical power construction document can be a historical power construction document, which can be used to construct a standard document. It is understood that there can be multiple first historical power construction documents. The more first historical power construction documents there are, the higher the accuracy of the standard document. A recognizable document is one in which the text is recognizable. It is understood that the first historical power construction document can be in image format; in this case, the text in the document is not recognizable but is an image, requiring single-character recognition of the image. The document type tag can be a tag used to characterize the document type. The document title can be any title in the document. The document content can be any content in the document, such as an event description, time, or other information, etc., and this embodiment of the invention does not impose any limitations on this. The content type of the document content can be the type of document content. For example, if the document content is text, the content type can be character type. If the document content is number, the content type can be numeric type. A standard document can be a document that can serve as a standard during the document verification process.

[0050] Specifically, the process involves acquiring a first historical power construction document, performing single-character recognition on the document to obtain an identifiable document, determining the document type tag of the identifiable document, acquiring the document title within the identifiable document, and acquiring the document content corresponding to the title. Semantic recognition is then performed on the document content to determine its content type. Based on the document type tag, document title, and document content type, a standard document is constructed. It should be noted that different document type tags can correspond to different standard documents. This embodiment of the invention does not limit the specific implementation method for determining the document type tag of the identifiable document, as long as the determination of the document type tag of the identifiable document can be achieved. Optionally, the standard document may include multiple attribute features. Each attribute feature can determine a column in the standard document to obtain a feature column, and each feature column can represent different document attribute features. For example, feature columns may include a title column, an event description column, or a solution column, etc.

[0051] S220. Input each of the first documents to be verified sequentially into a pre-trained document classification model to determine the document type of each of the first documents to be verified through the document classification model.

[0052] Optionally, before sequentially inputting each first document to be verified into the pre-trained document classification model, the process may further include: acquiring a set number of second historical power construction documents and sequentially inputting each second historical power construction document into the document classification model; extracting data from each second historical power construction document using the document classification model to obtain training document data; preprocessing the training document data using the document classification model to remove empty and abnormal data to obtain usable training data; extracting features and quantifying the usable training data using the document classification model to obtain usable training vectors; determining the text similarity between the usable training vectors and standard feature vectors using the document classification model; wherein the standard feature vectors are feature vectors determined based on standard documents; and adjusting the model parameters of the document classification model based on the text similarity.

[0053] The second historical power construction document can be a historical power construction document, which can be used to train the document classification model. The training document data can be any data from the second historical power construction document, such as the title, etc., and this embodiment of the invention does not impose any limitations on this. The training available data can be data obtained after preprocessing the training document data. The training available vector can be a feature vector obtained by converting the training available data into a vector format. The standard feature vector can be a feature vector obtained by converting data from a standard document into a vector format.

[0054] Specifically, before sequentially inputting each of the first documents to be verified into the pre-trained document classification model, a set number of second historical power construction documents can be obtained. These second historical power construction documents are then sequentially input into the document classification model. The model extracts training document data from each of the second historical power construction documents. This training document data is then preprocessed to remove empty and abnormal data, resulting in usable training data. Feature extraction and numerical processing are then performed on this usable training data to obtain usable training vectors. The document classification model then determines the text similarity between the usable training vectors and the standard feature vectors. Finally, the model parameters of the document classification model are adjusted based on the text similarity. Specifically, the standard feature vectors can be determined based on standard documents.

[0055] Optionally, determining the document type of each first document to be verified using a document classification model may include: determining the first document data of each first document to be verified using a document classification model; preprocessing the first document data using a document classification model to remove empty and abnormal data to obtain usable first document data; extracting features and quantifying the usable first document data using a document classification model to obtain document feature vectors; determining the text similarity between each document feature vector and a standard feature vector using a document classification model; wherein the standard feature vector is a feature vector determined based on a standard document; and determining the document type of each first document to be verified based on the text similarity using a document classification model.

[0056] The first document data can be data from a first document to be verified, such as the title data of the first document to be verified, and this embodiment of the invention does not impose any limitations on this. The first document usable data can be usable data obtained by preprocessing the first document data. The document feature vector can be a feature vector obtained by converting the first document usable data into a vector format.

[0057] Specifically, each first document to be verified is sequentially input into a pre-trained document classification model to determine the first document data of each first document to be verified. The first document data is then preprocessed to remove empty and abnormal data to obtain usable data. Feature extraction and numerical processing are then performed on the usable data to obtain document feature vectors. The text similarity between each document feature vector and the standard feature vector is then determined, and the document type of each first document to be verified is determined based on the text similarity.

[0058] Optionally, determining the document type of each first document to be verified based on text similarity may include: obtaining the document type label corresponding to the standard feature vector when the text similarity meets a set threshold; and determining the document type in the document type label as the document type of the first document to be verified.

[0059] Specifically, after determining the text similarity between each document's feature vector and the standard feature vector, it can be further determined whether the text similarity meets a set threshold. When the text similarity meets the set threshold, the document type in the document type label corresponding to the standard feature vector is determined as the document type of the first document to be verified. It can be understood that each standard feature vector can correspond to a document type label.

[0060] S230. Determine the target standard document in the standard documents according to the document type of each of the first documents to be verified; wherein the document type in the document type tag of the target standard document is the same as the document type of the first documents to be verified.

[0061] The target standard document can be a document of the same document type as the first document to be verified.

[0062] In this embodiment of the invention, after determining the document type of each first document to be verified through a document classification model, a target standard document in the standard documents can be further determined based on the document type of each first document to be verified. Specifically, the document type in the document type tag of the target standard document is the same as the document type of the first document to be verified.

[0063] S240. Determine the first document parameters to be verified for each of the first documents to be verified; wherein the first document parameters to be verified include document attribute features, document attribute feature order, and document content type corresponding to the document attribute features.

[0064] The first document parameter to be verified can be any parameter in the first document to be verified. The document attribute feature can be any document attribute in the first document to be verified, such as a file name or an event description; this embodiment of the invention does not impose any limitations on this. The order of the document attribute features can be the order of the document attribute features in the first document to be verified. The document content type can be the type of document content in the first document to be verified that corresponds to the document attribute feature.

[0065] In this embodiment of the invention, after determining the target standard document in the standard document according to the document type of each first document to be verified, the document attribute features, the order of the document attribute features, and the document content type corresponding to the document attribute features of each first document to be verified can be further determined.

[0066] S250. Determine the target standard document parameters of the target standard document; wherein, the target standard document parameters include target attribute features, target attribute feature order, and target content type corresponding to the target attribute features.

[0067] The target standard document parameters can be any parameter to be verified in the target standard document. The target attribute features can be any document attribute in the target standard document, such as a file name or an event description; this embodiment of the invention does not impose any limitations on this. The order of the target attribute features can be the order of the target attribute features in the target standard document. The target content type can be the type of document content in the target standard document corresponding to the target attribute features.

[0068] In this embodiment of the invention, after determining the document attribute features, the order of the document attribute features, and the document content type corresponding to the document attribute features of each first document to be verified, the target attribute features, the order of the target attribute features, and the target content type corresponding to the target attribute features of the target standard document can be further determined.

[0069] S260. If the parameters of the first document to be verified are the same as the parameters of the target standard document, determine that the first document to be verified meets the complete specification verification.

[0070] In this embodiment of the invention, after determining the target attribute features, the order of the target attribute features, and the target content type corresponding to the target attribute features of the target standard document, it can be further determined whether the parameters of the first document to be verified are the same as the parameters of the target standard document. If the parameters of the first document to be verified are the same as the parameters of the target standard document, it can be determined that the first document to be verified meets the complete specification verification. Specifically, if the document attribute features are the same as the target attribute features, the order of the document attribute features is the same as the order of the target attribute features, and the document content type is the same as the target content type, it can be determined that the parameters of the first document to be verified are the same as the parameters of the target standard document.

[0071] S270. Determine the items to be verified in each of the second documents to be verified, and combine the contents to be verified in each item to be verified into logical contents to be verified according to a preset document logical order.

[0072] The items to be verified can be feature items in the second document to be verified that require logical verification, such as time items or event items, etc., and this embodiment of the invention does not limit this. The content to be verified can be specific content in the items to be verified, such as specific time or specific information description, etc., and this embodiment of the invention does not limit this. The preset document logical order can be a pre-set logical order in which the content to be verified conforms when performing logical verification on the document. For example, assuming the second document to be verified is a project commencement application form, and the items to be verified are time items, including commencement time, project department review time, and construction unit approval time, then the preset document logical order can be commencement time -> project department review time -> construction unit approval time. The logical content to be verified can be the content to be verified composed of the content to be verified according to the preset document logical order.

[0073] In this embodiment of the invention, after determining that each first document to be verified meets the completeness specification verification, the items to be verified in each second document to be verified can be further determined, and the content to be verified in each item to be verified can be combined into logical content to be verified according to a preset document logical order. It is understood that the second document to be verified can be a document with a set logical relationship among the first documents to be verified that meets the completeness specification verification. It is also understood that the content to be verified in the logical content to be verified can be content from one second document to be verified, or content from multiple second documents to be verified; this embodiment of the invention does not impose any limitations on this.

[0074] S280. If it is determined that the logical content to be verified meets the preset logical rules, the second document to be verified is determined to meet the document logical verification.

[0075] The preset logical rules can be pre-defined logical rules. For example, assuming the document logical order is chronological logical order, the preset logical rules can be ascending order rules of time, etc., and this embodiment of the invention does not impose any limitations on this.

[0076] In this embodiment of the invention, after determining the items to be verified in each second document to be verified and assembling the contents to be verified in each item to be verified into logical contents to be verified according to a preset document logical order, it can be further determined whether the logical contents to be verified meet the preset logical rules, and when it is determined that the logical contents to be verified meet the preset logical rules, it is determined that the second document to be verified meets the document logical verification.

[0077] The technical solution of this embodiment acquires the power construction project documents to be verified, and identifies at least one first document to be verified and at least one second document to be verified from the power construction project documents. Each first document to be verified is sequentially input into a pre-trained document classification model to determine the document type of each first document to be verified. Based on the document type of each first document to be verified, a target standard document in the standard document is determined, and the first document parameters of each first document to be verified and the target standard document parameters of the target standard document are determined. If the parameters of the first document to be verified are the same as the parameters of the target standard document, it is determined that the first document to be verified meets the complete specification verification. This determines the items to be verified in each second document to be verified, and the content to be verified in each item to be verified is arranged into logical content to be verified according to a preset document logical order. If the logical content to be verified meets the preset logical rules, it is determined that the second document to be verified meets the document logical verification. This solves the problems of slow speed and low accuracy caused by manual verification of existing power construction documents, and can automatically verify power construction documents, improving the verification speed and accuracy of power construction documents.

[0078] To enable those skilled in the art to better understand the power construction document verification method of this embodiment, a specific example is used below for illustration. Figure 3 This is an example architecture diagram of a power construction document verification device provided in Embodiment 2 of the present invention, as shown below. Figure 3 As shown, the power construction document verification equipment may include an AI chip, speaker, storage chip, switching power supply, clock chip, buttons, LCD screen, double-sided automatic document feeder, and image sensor to realize the power construction document verification method.

[0079] Specifically, the main CPU and AI chip can utilize the Yizhi SH516 chip. The Yizhi SH516 chip employs advanced TSMC 22nm process technology, integrates 1Gb DDR3L memory, uses a QFN88 package, and has a small size of only 9*9mm, making it a highly integrated, low-power visual analysis platform. This chip integrates Yizhi's second-generation self-developed NPU, as well as a rich set of basic algorithms for people, vehicles, and objects, efficiently supporting intelligent vision applications such as object classification and recognition, text recognition, face detection and recognition, skeletal analysis, and gesture recognition. Image recognition can utilize the SC133GS chip from Entway Technology. The SC133GS chip is a 6.35mm (1 / 4”) diagonal CMOS active pixel type solid-state image sensor with a square pixel array (1080x1280) and 1.38M effective pixels. This chip features a variable global shutter capable of charge-integration time. It utilizes a triple power supply of analog 2.5V, digital 1.2V, and interface 1.8V, exhibiting low power consumption, high sensitivity, low dark current, and PLS characteristics. The memory chip can be the YMTC X2-9060 from Yangtze Memory Technologies Co., Ltd. (YMTC). The YMTC X2-9060 has 512GB of storage space and is based on… The YMTC X2-9060 utilizes Yangtze Memory Technologies' third-generation TLC 3D NAND flash memory chips with 2.0 technology. It boasts industry-leading storage density, I / O transfer speeds, and optimized power consumption, making it widely applicable in enterprise, embedded, and consumer storage products to meet the storage needs of various scenarios including cloud computing, big data, mobile devices, and personal consumer terminals. Due to the large volume of documents requiring review, manually placing each one was time-consuming and labor-intensive; therefore, an automated document feeder was installed to automatically deliver power construction documents.

[0080] Specifically, the power construction document verification method implemented through power construction document verification equipment may include the following:

[0081] (1) Collect Chinese characters and fonts from various types of historical power construction documents, and process these Chinese characters and semantics to form a commonly used Chinese character set and semantic set for power. Specifically, cut the single characters of various types of historical power construction documents, store the single character pictures, recognize the single character pictures, and then de-duplicate all the recognized Chinese characters. The de-duplicated Chinese characters are the commonly used Chinese character set for power. Exemplarily, the Chinese character set may include "pre", "budget", "electricity", "power", "usage", "user", etc. By extracting the content from the electronic document after single character recognition, extract the title and specific content (such as event description, date, and other information, etc.) from it, and summarize this information into a document table, with each column of the table corresponding to an attribute feature. Then perform natural language processing on this table document, analyze each feature column, and perform word segmentation on columns such as the title column, event description column, solution column, etc., so as to form a semantic set. Exemplarily, the semantic set may include "supervision", "planning", "construction drawing", "design change", "completion", "filing", "review", "transfer", "finalization", "construction", "test", "equipment", "winning the bid", "evaluation", "acceptance", etc.

[0082] (2)预置蚁群算法在人工智能芯片亿智SH516里面,并把汉字集与语义集固化到人工智能芯片里面。 (2) Preset the ant colony algorithm in the artificial intelligence chip EIZO SH516, and solidify the Chinese character set and semantic set into the artificial intelligence chip.

[0083] (3) Before verifying the power construction document, it is necessary to scan and recognize the power construction document. The main MCU EIZO SH516 notifies the automatic document feeder to convey the power construction document to be scanned to the recognition area. When the recognition area detects a document, the image sensor SC133GS starts to automatically focus and scan the power construction document to be scanned, then preprocess the scanned image (such as grayscale processing), and transfer the scanned image to the NPU of EIZO SH516 through the data bus for recognition, and display it on the liquid crystal screen. For documents with a low image recognition rate, prompt to refocus and scan again. Due to problems such as printers, paper, ink, etc., the printed fonts of the power construction document to be scanned may have various defects. During the image acquisition process, the placement position, the pitch angle of the lens, the light during acquisition, etc. will also affect the quality of the acquired image, resulting in font defects, such as unclear fonts, tilted fonts, and missing parts in the font strokes, etc. These will seriously affect the subsequent recognition effect. In order to reduce recognition errors, the defective parts can be corrected on the liquid crystal screen, and at the same time, the corrected results can be sent to the Chinese character set for the program to learn in real time.

[0084] (4) After successful text recognition preprocessing, the collected dataset is subjected to text recognition and semantic association analysis and label processing. The dataset is divided into training set and test set, and a document recognition feature set is constructed to detect and recognize the integrity and standardization of construction documents. Then, the training set in the dataset is used for recognition training and parameter adjustment. After training, the recognition program is solidified into the AI ​​chip Yizhi SH516 to verify the integrity, standardization and logical relationship between the documents, which can greatly improve the cost of manual inspection of construction documents.

[0085] Taking training data as an example, if we need to train on titles, the titles in all the collected power construction document datasets would serve as the training data, which can be divided into training and testing sets. This attribute is then preprocessed, including word segmentation, stop word removal, and TF-IDF numerical conversion, transforming it into labels or vectors. Next, cosine similarity is used to calculate the similarity between two sets of data or two titles, with the loss rate being 1 - similarity. By training with these samples and labels, when the model receives a new construction document, it first preprocesses and performs word segmentation and vectorization, comparing the similarity between this new document and historical documents. If the similarity is greater than a certain threshold, the new document and historical documents are considered to belong to the same category; if it is lower than the threshold, a new category is created and stored. For example, various types of power construction documents, such as construction drawing review records, construction drawing design budgets, as-built drawing preparation instructions, comparison tables of design changes and as-built drawing modifications, production project archive handover records, and production project document filing review forms, are analyzed and processed as training data. The table titles are then extracted, and the above training data method is used for training to obtain labels for each type of title. Then, the similarity between these data and new documents is calculated. The higher the similarity, the stronger the association between the two documents, and the new document is classified into the category with the highest similarity.

[0086] Document integrity and standardization verification: For example, assuming a document contains main content such as title, event description, and solution, identify whether the new document contains these attribute columns. Extract the corresponding attribute values ​​from the new document's table based on each attribute column. A pre-set scoring standard (e.g., 80%) is used. When the attribute values ​​meet the pre-set scoring standard compared to the original document, the document is considered complete. Document standardization can be achieved by maintaining the same format as the original document. Based on the characteristics of each attribute column, check the order of these characteristics in both documents row by row. If the order of characteristics in the document to be verified is not significantly different from the order of characteristics in the historical document, the document is considered standard. For example, in analyzing the integrity and standardization of an engineering settlement audit finalization form document, when obtaining the document to be verified, compare the three attributes: file name, project name, and final settlement construction cost. If all three attributes are present, the document to be verified meets the integrity requirement. Then, compare the order of these three attributes. If the attribute order is the same in both documents, the document to be verified meets the standardization requirement.

[0087] Determining the logical relationships between sentences can be achieved through text normalization, removal of special characters and stop words, stemming, understanding text syntax and structure, part-of-speech tagging, shallow analysis, dependency-based analysis, and component structure analysis. The results will reveal the logical relationships. For example, analyzing the logical relationships within a project commencement application form involves text normalization, stemming, and deep learning to understand the text's syntax and structure. Sentences are broken down into subject, verb, and object, and the relationships between words are analyzed. Based on dependency relationships, verbs are considered the root of the sentence, and all other words are directly or indirectly connected to the root verb, indicating a relationship. Component structure analysis identifies the constituent parts of a sentence, its key purpose being to uncover their internal structure and relationships, thus revealing the logical relationships between sentences. For example, in a project commencement application form, normally the signature date must precede the commencement date. The signature dates can be equal from top to bottom or gradually increase, with the last signature date being later than the commencement date. In another example, the content of the document needs to be determined based on the project content. For example, a software project does not require site leveling or major materials and equipment.

[0088] The above technical solution can automatically verify power construction documents, improving the verification speed and accuracy.

[0089] Example 3

[0090] Figure 4 This is a schematic diagram of a power construction document verification device provided in Embodiment 3 of the present invention, as shown below. Figure 4As shown, the device includes: a document determination module 410, a document type determination module 420, a completeness and specification verification module 430, and a document logic verification module 440, wherein:

[0091] The document to be verified determination module 410 is used to obtain the power construction project documents to be verified, and determine at least one first document to be verified and at least one second document to be verified among the power construction project documents to be verified; wherein, the second document to be verified is a document with a set logical relationship among the first documents to be verified.

[0092] The document type determination module 420 is used to sequentially input each of the first documents to be verified into a pre-trained document classification model, so as to determine the document type of each of the first documents to be verified through the document classification model;

[0093] The complete specification verification module 430 is used to perform complete specification verification on each of the first documents to be verified according to the document type of each of the first documents to be verified.

[0094] The document logic verification module 440 is used to perform document logic verification on each of the second documents to be verified, provided that the complete specification verification results of each of the first documents to be verified meet the complete specification verification conditions.

[0095] The technical solution of this embodiment obtains the power construction project documents to be verified, identifies the first and second documents to be verified, and sequentially inputs each first document to be verified into a pre-trained document classification model. The document classification model determines the document type of each first document to be verified, and performs a complete specification verification on each first document to be verified based on its document type. Then, if the complete specification verification results of each first document to be verified meet the complete specification verification conditions, a document logic verification is performed on each second document to be verified. This solves the problems of slow speed and low accuracy caused by manual verification of existing power construction documents, and can automatically verify power construction documents, improving the verification speed and accuracy.

[0096] Optionally, the document determination module 410 can be specifically used to: obtain a first historical power construction document, and perform single-character recognition on the first historical power construction document to obtain an identifiable document; determine the document type tag of the identifiable document; obtain the document title in the identifiable document; obtain the document content corresponding to the document title, and perform semantic recognition on the document content to determine the content type of the document content; and construct a standard document based on the document type tag, document title, and content type of the document content.

[0097] Optionally, the document type determination module 420 can be specifically used for: acquiring a set number of second historical power construction documents and sequentially inputting each second historical power construction document into the document classification model; extracting data from each second historical power construction document through the document classification model to obtain training document data; preprocessing the training document data through the document classification model to remove empty and abnormal data to obtain usable training data; extracting features and quantifying the usable training data through the document classification model to obtain usable training vectors; determining the text similarity between the usable training vectors and the standard feature vectors through the document classification model; wherein the standard feature vectors are feature vectors determined based on standard documents; and adjusting the model parameters of the document classification model based on the text similarity.

[0098] Optionally, the document type determination module 420 can be specifically used to: determine the first document data of each first document to be verified through a document classification model; preprocess each first document data through a document classification model to remove empty and abnormal data to obtain usable data of the first document; extract features and perform numerical processing on the usable data of the first document through a document classification model to obtain document feature vectors; determine the text similarity between each document feature vector and the standard feature vector through a document classification model; wherein the standard feature vector is a feature vector determined based on a standard document; and determine the document type of each first document to be verified based on the text similarity through a document classification model.

[0099] Optionally, the document type determination module 420 can be further used to: obtain the document type label corresponding to the standard feature vector when the text similarity meets the set threshold; and determine the document type in the document type label as the document type of the first document to be verified.

[0100] Optionally, the complete specification verification module 430 can be specifically used to: determine the target standard document in the standard documents based on the document type of each first document to be verified; wherein the document type in the document type tag of the target standard document is the same as the document type of the first document to be verified; determine the first document to be verified parameters of each first document to be verified; wherein the first document to be verified parameters include document attribute features, document attribute feature order, and document content type corresponding to the document attribute features; determine the target standard document parameters of the target standard document; wherein the target standard document parameters include target attribute features, target attribute feature order, and target content type corresponding to the target attribute features. If the first document to be verified parameters are the same as the target standard document parameters, determine that the first document to be verified meets the complete specification verification.

[0101] Optionally, the document logic verification module 440 can be specifically used to: determine the items to be verified in each second document to be verified, and assemble the contents to be verified in each item to be verified into logical contents to be verified according to a preset document logic order; and determine that the second document to be verified satisfies the document logic verification if the logical contents to be verified satisfy the preset logic rules.

[0102] The power construction document verification device provided in this embodiment of the invention can execute the power construction document verification method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0103] Example 4

[0104] Figure 5 A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention 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 processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0105] like Figure 5 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 may also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0106] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0107] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the power construction document verification method.

[0108] In some embodiments, the power construction document verification method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the power construction document verification method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the power construction document verification method by any other suitable means (e.g., by means of firmware).

[0109] Various embodiments 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-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0110] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0111] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0112] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device 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 provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, 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 sound input, voice input, or tactile input).

[0113] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0114] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0115] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0116] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for verifying power construction documents, characterized in that, include: Obtain the power construction project documents to be verified, and determine at least one first document to be verified and at least one second document to be verified from the power construction project documents to be verified; wherein, the second document to be verified is a document with a set logical relationship among the first documents to be verified; Each of the first documents to be verified is sequentially input into a pre-trained document classification model to determine the document type of each of the first documents to be verified through the document classification model. Based on the document type of each of the first documents to be verified, a complete and standardized verification is performed on each of the first documents to be verified. If the complete specification verification results of each of the first documents to be verified meet the complete specification verification conditions, document logic verification is performed on each of the second documents to be verified. Prior to obtaining the power construction project document to be verified, the process also includes: Obtain the first historical power construction document and perform single-character recognition on the first historical power construction document to obtain a recognizable document; Determine the document type tag of the identifiable document; Obtain the document title from the identifiable document; Obtain the document content corresponding to the document title, and perform semantic recognition on the document content to determine the content type of the document content; A standard document is constructed based on the document type tag, the document title, and the content type of the document content; wherein different document type tags correspond to different standard documents; The step of performing a complete specification verification on each of the first documents to be verified according to their document types includes: Based on the document type of each of the first documents to be verified, a target standard document is determined in the standard documents; wherein the document type in the document type tag of the target standard document is the same as the document type of the first documents to be verified. Determine the first document parameters for each of the first documents to be verified; wherein, the first document parameters include document attribute features, the order of document attribute features, and the document content type corresponding to the document attribute features; wherein, the document attribute features include at least a file name and an event description; the document content type is the type of document content in the first document to be verified that corresponds to the document attribute features; Determine the target standard document parameters; wherein, the target standard document parameters include target attribute features, the order of target attribute features, and the target content type corresponding to the target attribute features; If the parameters of the first document to be verified are the same as the parameters of the target standard document, it is determined that the first document to be verified meets the complete specification verification.

2. The method according to claim 1, characterized in that, Before sequentially inputting each of the first documents to be verified into the pre-trained document classification model, the method further includes: Obtain a set number of second historical power construction documents, and input each of the second historical power construction documents into the document classification model in sequence; Using the document classification model, data is extracted from each of the second historical power construction documents to obtain training document data. The document classification model is used to preprocess the training document data to remove empty and abnormal data, thereby obtaining usable training data. The document classification model is used to extract features and quantify the available data to be trained to obtain the available vector to be trained. The document classification model is used to determine the text similarity between the trainable vector and the standard feature vector; wherein the standard feature vector is a feature vector determined based on the standard document. The model parameters of the document classification model are adjusted based on the text similarity.

3. The method according to claim 1, characterized in that, The step of determining the document type of each of the first documents to be verified through the document classification model includes: The first document data for each of the first documents to be verified is determined using the document classification model. The document classification model is used to preprocess each of the first document data to remove empty and abnormal data from the first document data, thereby obtaining usable data of the first document. Using the document classification model, feature extraction and numerical processing are performed on the available data of the first document to obtain a document feature vector. The document classification model is used to determine the text similarity between each document feature vector and a standard feature vector; wherein the standard feature vector is a feature vector determined based on the standard document. The document type of each of the first documents to be verified is determined based on the text similarity using the document classification model.

4. The method according to claim 3, characterized in that, The step of determining the document type of each of the first documents to be verified based on the text similarity includes: If the text similarity meets a set threshold, obtain the document type label corresponding to the standard feature vector; The document type in the document type label is determined as the document type of the first document to be verified.

5. The method according to claim 1, characterized in that, The document logic verification of each of the second documents to be verified includes: Determine the items to be verified in each of the second documents to be verified, and combine the contents to be verified in each item to be verified into logical contents to be verified according to a preset document logical order; If the logical content to be verified is determined to meet the preset logical rules, the second document to be verified is determined to meet the document logical verification.

6. A power construction document verification device, characterized in that, include: The document to be verified module is used to obtain power construction project documents to be verified, and to determine at least one first document to be verified and at least one second document to be verified from the power construction project documents to be verified; wherein, the second document to be verified is a document with a set logical relationship among the first documents to be verified. The document type determination module is used to sequentially input each of the first documents to be verified into a pre-trained document classification model, so as to determine the document type of each of the first documents to be verified through the document classification model; The complete specification verification module is used to perform complete specification verification on each of the first documents to be verified according to the document type of each document to be verified. The document logic verification module is used to perform document logic verification on each of the second documents to be verified, provided that the complete specification verification results of each of the first documents to be verified meet the complete specification verification conditions. The document verification module is specifically used for: The process involves: acquiring a first historical power construction document and performing single-character recognition on the document to obtain an identifiable document; determining the document type tag of the identifiable document; acquiring the document title in the identifiable document; acquiring the document content corresponding to the document title and performing semantic recognition on the document content to determine the content type of the document content; and constructing a standard document based on the document type tag, the document title, and the content type of the document content, wherein different document type tags correspond to different standard documents. Specifically, the complete specification verification module is used for: Based on the document type of each of the first documents to be verified, a target standard document in the standard documents is determined; wherein the document type in the document type tag of the target standard document is the same as the document type of the first documents to be verified; a first document parameter to be verified is determined for each of the first documents to be verified; wherein the first document parameter to be verified includes document attribute features, document attribute feature order, and document content type corresponding to the document attribute features; wherein the document attribute features include at least file name and event description; the document content type is the type of document content in the first documents to be verified that corresponds to the document attribute features; a target standard document parameter to the target standard document is determined; wherein the target standard document parameter includes target attribute features, target attribute feature order, and target content type corresponding to the target attribute features; if the first document parameter to be verified is the same as the target standard document parameter, the first document to be verified is determined to meet the complete specification verification.

7. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the power construction document verification method according to any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the power construction document verification method according to any one of claims 1-5.

Images