Production project whole-process transparent management and control index analysis method, system, equipment and medium

Abstract

The invention relates to the technical field of production project management and control, and discloses a production project whole-process transparent management and control index analysis method, system and device and a medium, and the method comprises the steps: obtaining multi-source data, and integrating the multi-source data to obtain first index data; constructing a digital production project management model according to the first index data, and checking production project links by using the digital production project management model to obtain a first abnormal data list; based on the first abnormal data list, analyzing the first index data by utilizing a first natural language processing method in combination with a first rule to obtain second abnormal data; and alarm information is generated according to the second abnormal data, and a first notification mechanism is triggered to notify related personnel to take management and control measures, so that a large amount of complex and tedious work such as traditional data collection, arrangement and checking is avoided, and a large amount of human resources are saved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of production project management and control, and in particular to a production project whole-process transparent management and control index analysis method, system, device and medium. BACKGROUND

[0002] In the related index analysis work of the whole-process transparent management of the project, the data sources are more dimensional, and it is necessary to regularly evaluate and analyze to output daily, weekly and monthly reports. The related data drilling relies on manual work, and the repeatability and singleness are strong, the technical content is not high, the evaluation workload of the evaluation personnel is large, the evaluation personnel cannot focus on the key links of data anomalies and cannot take timely disposal measures for the abnormal links.

[0003] The method of the present application aims to liberate the mechanical work of data drilling and can perform preliminary analysis according to the instructions of the evaluation personnel, greatly improving the work efficiency and eliminating the problems of data omission and data checking complexity. The average number of construction operation plans related to production of external construction units in the province per day is hundreds, and the related operation plans involve the establishment and association of production projects according to the requirements of the company. The operation plan involves the operation section and content of the key line and key equipment, and whether the content is consistent with the establishment content and whether there is the risk of unauthorized expansion of standards needs to be checked by manual work, which is time-consuming and easy to miss. This project aims to solve the work of pure manual screening. According to the pre-set key instructions and rules, the irrelevant content of the operation plan and the project content is preliminarily judged, the alarm mechanism is triggered and delivered to the evaluation personnel or the relevant person in charge, and timely measures are taken to supervise the closed loop. This helps the evaluation personnel to control the related risks of false establishment and false acceptance in real time every day, and improves the disposal efficiency. SUMMARY

[0004] In view of the above existing problems, the present application is proposed.

[0005] Therefore, the present application provides a production project whole-process transparent management and control index analysis method, system, device and medium to solve the problems of multiple data channels required in the transparent management and control process of production projects, low work efficiency of mechanical data drilling, data omission, data checking complexity, passive manual screening of "four false risks", inability to grasp the risk situation in time, and inability to supervise the closed loop in time.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a production project whole-process transparent management and control index analysis method, comprising:

[0008] Obtaining multi-source data, and obtaining first index data by integrating the multi-source data;

[0009] construct a digital production project management model according to the first index data, check the production project link by using the digital production project management model, and obtain a first abnormal data list;

[0010] Based on the first abnormal data list, the first index data is analyzed by using a first natural language processing method combined with a first rule to obtain second abnormal data.

[0011] According to the second abnormal data, an alarm information is generated, and a first notification mechanism is triggered to notify relevant personnel to take control measures.

[0012] As a preferred scheme of the production project whole-process transparent control index analysis method, wherein:

[0013] The multi-source data is matched by using a first matching method to obtain key information.

[0014] The indexes in the key information are sorted and valued to obtain the first index data for storage.

[0015] As a preferred scheme of the production project whole-process transparent control index analysis method, wherein:

[0016] According to a preset rule, each entity and corresponding relationship in the digital knowledge graph model are checked.

[0017] If the entity or the corresponding relationship does not conform to the preset rule, the first abnormal data is recorded, and a first abnormal data list is generated.

[0018] The beneficial effect of the preferred technical scheme is that by checking the entity relationship, the abnormality is efficiently found, and the control accuracy is improved.

[0019] As a preferred scheme of the production project whole-process transparent control index analysis method, wherein:

[0020] The first index data is scanned and grabbed by using a first natural language processing method.

[0021] The first index data is converted and analyzed according to a first rule by using keyword matching to obtain second abnormal data.

[0022] The beneficial effect of the preferred technical scheme is that by natural language processing and keyword matching, the abnormal data is quickly identified, and the analysis efficiency and accuracy are improved.

[0023] As a preferred solution of the method for analyzing indicators of transparent control over the entire production process of the present invention, generating alarm information according to the second abnormal data includes:

[0024] Define risk levels, warning levels, and the corresponding relationships between risk levels and warning levels;

[0025] Perform risk assessment on the second abnormal data according to the risk level to obtain a risk assessment result;

[0026] According to the risk assessment results, the corresponding level of alarm is triggered and the alarm information is obtained.

[0027] The beneficial effects of this preferred technical solution are that it can improve risk management capabilities by accurately assessing risks, triggering alarms at different levels, and responding to anomalies in a timely manner.

[0028] As a preferred solution of the method for analyzing indicators of transparent control over the entire production process of the present invention, matching multi-source data using the first matching method includes:

[0029] Performing string matching and feature code verification on the multi-source data to obtain first multi-source data;

[0030] Preprocessing the first multi-source data to obtain key information;

[0031] Build a keyword library for key information of analytical indicators, match keywords, and extract entity information elements;

[0032] Use word embedding to convert the text into a vector representation, train the semantic matching model, and obtain the first semantic matching model;

[0033] Calculating the similarity between the searched text to be spliced ​​and the entity information element using the first semantic matching model;

[0034] Determine the similarity between two entities;

[0035] If the similarity is greater than the first threshold, the entities are identical, and the corresponding list records are concatenated.

[0036] The beneficial effect of this preferred technical solution is that it achieves efficient data integration and anomaly identification through matching and semantic analysis of multi-source data, thereby improving the accuracy and efficiency of data processing.

[0037] As a preferred solution of the method for analyzing indicators of transparent control over the entire production process of the present invention, it also includes:

[0038] Convert multi-source data into a unified feature vector;

[0039] The keywords are taken as nodes, a correlation subgraph is constructed, and the subgraph is mapped with a relationship node in a relationship network, so as to realize correlation of multi-source data.

[0040] For nodes with similar semantics, the correlation between multi-source data is obtained by fusing through calculation of the shortest semantic distance.

[0041] In a second aspect, the present application provides a production project whole-process transparent management and control index analysis system, comprising:

[0042] A data integration module is configured to acquire multi-source data, and obtain first index data by integrating the multi-source data.

[0043] A checking module is configured to construct a digital production project management model according to the first index data, and check a production project link by using the digital production project management model to obtain a first abnormal data list.

[0044] A data analysis module is configured to analyze the first index data by using a first natural language processing method combined with a first rule based on the first abnormal data list to obtain second abnormal data.

[0045] An alarm information generation module is configured to generate alarm information according to the second abnormal data, and trigger a first notification mechanism to notify relevant personnel to take management and control measures.

[0046] In a third aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and wherein the processor implements the steps of the production project whole-process transparent management and control index analysis method when executing the computer program.

[0047] In a fourth aspect, the present application provides a computer readable storage medium storing a computer program, wherein the computer program implements the steps of the production project whole-process transparent management and control index analysis method when executed by a processor.

[0048] Compared with the prior art, the present application has the following beneficial effects: the present application has intelligent perception and closed-loop management and control technology for "four virtual risks" of a production project through multi-source data acquisition, integration and analysis capability of the production project based on intelligent technology; through application of the present application, work efficiency is greatly improved, a large amount of complex and tedious work such as traditional data collection, sorting and checking is avoided, a large amount of human resources is saved, work efficiency is greatly improved, the focus is shifted to the decision-making level, and the production project management and control capability is obviously improved. BRIEF DESCRIPTION OF DRAWINGS

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without any creative effort on the basis of these drawings.

[0050] Figure 1 The overall flow logic diagram of the production project whole-process transparent management and control index analysis method provided by an embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0051] In order to make the above-mentioned purposes, features and advantages of the present application more apparent and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the drawings in the specification. Obviously, the described embodiments are only some embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without any creative effort should be within the protection scope of the present application.

[0052] Embodiment 1, refer to Figure 1 Table 1, a production project whole-process transparent management and control index analysis method is provided by an embodiment of the present application, which comprises:

[0053] S100: Obtain multi-source data, and obtain first index data by integrating the multi-source data;

[0054] S200: Construct a digital production project management model according to the first index data, and check the production project link by using the digital production project management model to obtain a first abnormal data list;

[0055] S300: Based on the first abnormal data list, analyze the first index data by using a first natural language processing method combined with a first rule to obtain second abnormal data;

[0056] S400: Generate an alarm information according to the second abnormal data, and trigger a first notification mechanism to notify relevant personnel to take control measures.

[0057] It should be noted that by multi-source data integration, digital model checking, natural language processing analysis and intelligent alarm notification, the transparent management and control of the whole process of the production project is realized. The abnormal data can be efficiently identified, the risk can be accurately assessed and the relevant personnel can be timely notified to take measures, which greatly improves the efficiency and accuracy of the production project management, reduces the tedious work of manual checking, reduces the risk of omission and error, and significantly enhances the project control capability.

[0058] In the embodiments of the present application, the above step S100 comprises the following sub-steps A1-A2:

[0059] In A1: the multi-source data is matched by using a first matching method, to obtain key information;

[0060] In A2: the indexes in the key information are sorted and valued, to obtain first index data for storage;

[0061] Specifically, the robot process automation (RPA) is used to realize the automatic collection and integration of cross-system and multi-format data by simulating manual operation;

[0062] The RPA robot simulates manual data browsing, captures and caches the browsed data, and in the process of data collection, the pre-set keywords and feature codes are used to match the data collected by the RPA robot, the data items that hit the features are stored and downloaded, and the indexes are constructed according to the keywords to form index data for storage;

[0063] The information such as project commencement, milestone plan, completion, fund plan and image progress of the power grid management platform is obtained; the information such as project operation correlation and complete rate of engineering quantity certification of the transparent system is obtained; the key information such as time and amount in the data is used to sort and value the indexes.

[0064] In the embodiment of the present application, after the A1-A2 steps in the above step S100 are completed, the following steps A3-A12 are further included;

[0065] In A3: string matching and feature code checking are performed on the multi-source data, to obtain first multi-source data;

[0066] In A4: the first multi-source data is preprocessed to obtain key information;

[0067] In A5: a keyword library of analysis index key information is constructed, the keywords are matched, and entity information elements are extracted;

[0068] In A6: the text is converted into a vector representation by using word embedding, a semantic matching model is trained, and a first semantic matching model is obtained;

[0069] In A7: the similarity between the search text to be spliced and the entity information elements is calculated by using the first semantic matching model;

[0070] In A8: the similarity between two entities is judged;

[0071] In A9: if the similarity is greater than a first threshold, the entities are the same, and the corresponding list records are spliced;

[0072] In A10: the multi-source data is converted into a unified feature vector;

[0073] In A11: the keywords are taken as nodes, an association subgraph is constructed, and the subgraph is mapped with the relationship nodes in the relationship network, so as to realize the association of multi-source data;

[0074] In A12: for nodes with similar semantics, the shortest semantic distance is calculated for fusion, so as to obtain the association relationship between multi-source data.

[0075] The collected multi-source data are digitally converted by using a semantic matching technology, and are integrated into a basic model with correlation and digitalization for production project transparent control, so as to form a unified index by associating multi-source data, and the heterogeneous data need to be matched;

[0076] In an optional embodiment, the first matching method can be a machine learning-based classifier matching, a large amount of labeled sample data is obtained, key features are extracted from the sample data, a suitable machine learning algorithm is selected, the sample data is trained, a classifier model is generated, the data to be matched is input into the classifier, it is judged whether the data conforms to the preset rule, and a matching result is output;

[0077] In an optional embodiment, the first matching method can be a dynamic matching based on a rule engine, a rule is defined, the defined rule is loaded into the rule engine, the data to be matched is input into the rule engine, the engine matches the data one by one according to the preset rule, and the rule engine outputs a matching result;

[0078] In the embodiment of the application, the first matching method includes combining traditional rule matching based on preset keywords and feature codes with matching based on semantic analysis technology;

[0079] Specifically, the traditional rule matching includes accurate string matching and feature code checking.

[0080] The accurate string matching is to determine the consistency of data by detecting whether a field completely matches a preset value, for example, detecting whether the "planned completion construction time" field completely matches a preset date format; a regular expression is used to define the format rule of the field, such as the completion time: (0[1-9]|1[0-2])[-\ / ](0[1-9]|\d|3)[-\ / ]\d{4}$ matches the completion time: 2025-01-01 or 2025 / 01 / 01 and other variants.

[0081] The feature code checking is to verify the legality of data by detecting a specific feature code in the data, for example, detecting a 16-bit digital feature code, which can quickly find and check data by calculating the hash value of the feature code.

[0082] The first preprocessing of the data based on the semantic analysis technology includes removing irrelevant characters, special symbols and stop words in the text to improve the accuracy of subsequent processing; and using a Jieba Chinese word segmentation tool to divide the text into independent lexical units;

[0083] According to the business requirements, a keyword library of analysis index key information is constructed, the word frequency in the keyword library is increased, and is loaded into the Jieba word segmentation tool dictionary to improve the accuracy of word segmentation, so as to determine whether the text contains relevant information. When Chinese text is segmented, it is usually according to the position, and each character is marked using four-tag. Four-tag is four possible states of each character in a word, B, M, E, S, as shown in Table 1:

[0084] Table 1 Four possible states of each character in a word

[0085] Wordpiece tags Tag explanation B Begin - This word is at the beginning of a wordpiece M Middle - This word is in the middle of a wordpiece E End - This word is at the end of a wordpiece S Single - This word is a single wordpiece

[0086] Extracting entity information elements from the text, in the present application, mainly the name of the power transformation, named entity recognition is the basis for solving many natural language processing problems, and is also the most basic task of entity extraction;

[0087] Using a word embedding method to convert the text into a vector representation, and using a labeled data set to train a semantic matching model, obtaining a first semantic matching model as a trained semantic matching model, using the trained semantic matching model to calculate the similarity between the search text to be spliced and the power transformation name; After the model processing, the text to be spliced obtains a vector representation, and the power transformation name obtains another vector representation after the same processing, and then the similarity between the two vectors is calculated to obtain the final similarity score. The vector similarity is represented as:

[0088]

[0089] Wherein, v and w represent two vectors respectively, and ||v|| and ||w|| represent the norm of vectors v and w respectively;

[0090] The first threshold is determined according to the specific application scenario, data characteristics and business requirements. The higher the similarity, the more similar the two entities are. When the similarity is greater than the first threshold, the two entities are the same, and the corresponding two list records are spliced;

[0091] According to the predetermined setting, the project name extracted by RPA is stored in a structured form to form a project library information, and according to the subsequent analysis logic, the progress, funds, plans and other completion conditions of the project are calculated and analyzed.

[0092] It should be noted that by combining traditional rule matching and matching method based on semantic analysis technology, efficient and accurate matching of multi-source data is realized; the traditional method ensures the accuracy of data format and feature code, and the semantic analysis makes up for the deficiency at the lexical level and can handle synonyms, phrase rearrangement and other problems; in addition, through the RPA technology, the project name is extracted and stored in a structured manner, which further improves the data processing efficiency and accuracy, and provides reliable support for subsequent project progress, fund, plan and other analysis.

[0093] In the embodiment of the present application, the above step S200 comprises the following sub-steps B1-B2;

[0094] In B1: according to the preset rules, each entity and corresponding relationship in the digital knowledge graph model is checked;

[0095] In B2: if the entity or the corresponding relationship does not conform to the preset rules, the first abnormal data is recorded, and a first abnormal data list is generated;

[0096] Specifically, the digital knowledge graph model is customized and preset logic according to actual needs in the program development process, which realizes the full traversal and automatic checking of the progress or changes of each link in the project process according to the pre-defined rules, and automatically gives the checking process and the result list of the existing abnormalities. One-key confirmation of the user eliminates the traditional data collection, sorting, checking and other complex and tedious work, greatly improves the work efficiency, shifts the focus to the decision-making level, and significantly improves the production project control ability.

[0097] In an optional embodiment, the preset rules can be quality control rules to check whether the project documents such as design files, acceptance reports and the like conform to the preset format and content requirements. For example, the design file must contain key parts such as project background, design target, technical scheme, etc.; the acceptance report must clearly list the acceptance standards, test results and the like; if the document lacks key parts or the format does not meet the requirements, it is marked as abnormal;

[0098] In an optional embodiment, the preset rules can be resource management rules to check whether the personnel configuration of the project team is reasonable and meets the project progress and task requirements. For example, according to the workload and difficulty of the project task, the number of professional personnel required for each task and the skill requirements are preset; if it is found that the personnel configuration of a certain task is insufficient or the skills are not matched, it is marked as abnormal and personnel adjustment is suggested;

[0099] In the embodiment of the present application, the preset rules include: checking whether the key fields in the project data conform to the preset format, such as the date field must conform to a specific format; checking the data integrity to ensure that the key information such as the project name, the key equipment name and the like is not missing;

[0100] Check whether the project is advancing according to the planned milestone plan, such as the preliminary design must be completed within 30 days after the project starts, otherwise it is marked as an exception; check the association between tasks to ensure that the previous task is completed before starting the subsequent task;

[0101] Check whether the project has "four false risks" such as false establishment, false acceptance, and unauthorized expansion of standards, such as when the key equipment operation section in the work plan does not match the establishment content, it is marked as an exception; define risk levels according to project amount, delay length and other indicators, such as when the project amount exceeds 100,000 yuan and the delay exceeds 30 days, it is marked as high risk;

[0102] It should be noted that through automatic checking and anomaly detection, tedious data collection and sorting work is eliminated, users can confirm the results with one key, work efficiency is improved, decision-making focus is assisted, and production project control ability is significantly enhanced.

[0103] In the embodiment of the application, the above step S300 includes the following sub-steps C1-C2;

[0104] In C1, the first natural language processing method is used to scan and capture the first index data;

[0105] In C2, the keywords are matched, the first index data is converted and analyzed according to the first rule, and the second abnormal data is obtained.

[0106] In an optional embodiment, the first natural language processing method can be named entity recognition, and a pre-trained named entity recognition model is used to scan the daily generated work plan and work content text, and identify key entities such as project name, device name, and work site; according to business requirements, define specific named entity rules, for example, for entities such as "substation name" and "line name" in power projects, supplement the recognition through regular expressions or keyword matching;

[0107] In an optional embodiment, the first natural language processing method can be text classification, a large amount of annotated work plan text data is collected, including normal work plans and abnormal work plans with "four false risks", and a text classification model is trained using these data, key features are extracted from the text, the text is converted into a feature vector, and input into the classification model, the daily generated work plan text is classified in real time, and it is judged whether it belongs to the abnormal category; if the model predicts an exception, further analyze the specific risk type;

[0108] In the embodiment of the application, the first natural language processing method includes semantic analysis technology;

[0109] Specifically, by using semantic analysis and other technologies, the daily generated work plans, work contents, associated projects and the like are automatically scanned and captured, semantic matching is performed by using keywords, the data is structured converted according to the first rule, and intelligent analysis is performed based on the first rule, and abnormal information of four virtual risks such as unauthorized expansion of standards and expansion of construction scale in the project is actively discovered;

[0110] In an optional embodiment, the first rule can be a compliance rule, and the project construction process must comply with environmental protection requirements such as noise emission and waste disposal; if a behavior that does not comply with the environmental protection standards is detected, it is marked as an exception;

[0111] In an optional embodiment, the first rule can be a progress optimization rule, and priorities are assigned according to the importance and urgency of tasks; high-priority tasks must be completed first, and if a high-priority task is delayed, it is marked as an exception;

[0112] In the embodiment of the application, the first rule includes that the date field must comply with a specific format, the project number must be 16 digits, the fund plan field must be a positive number; the project name field must match the preset keyword; and the key equipment name must match the name in the known equipment list.

[0113] It should be noted that by using semantic analysis technology to automatically capture and analyze work plan data, keyword matching and rule analysis are used to efficiently identify "four virtual risk" abnormal information, improve risk control efficiency and accuracy, and help project compliance promotion.

[0114] In the embodiment of the application, the above step S400 includes the following sub-steps D1-D3.

[0115] In D1, the correspondence between the risk level, the alarm level, the risk level and the alarm level is defined.

[0116] In D2, the second abnormal data is risk evaluated according to the risk level, and a risk evaluation result is obtained.

[0117] In D3, according to the risk evaluation result, an alarm of a corresponding level is triggered, and alarm information is obtained.

[0118] Specifically, according to the preset risk grading standard, the system grades the identified risks. The risk grading standard can be based on multiple indicators such as project amount, delay duration (proportion of total construction period), task completion rate, etc.

[0119] For example, if the project amount is less than 500,000 yuan, the delay duration is less than 10% of the total construction period, and it is low risk; if the project amount is between 500,000 yuan and 1,000,000 yuan, the delay duration is between 10% and 20% of the total construction period, and it is medium risk; if the project amount is more than 1,000,000 yuan, the delay duration is more than 20% of the total construction period, and it is high risk;

[0120] The corresponding relationship between the risk level and the alarm level is as follows:

[0121] Low risk corresponds to low alarm level, and the first notification mechanism is ordinary mail or message notification;

[0122] Medium risk corresponds to medium alarm level, and the first notification mechanism is urgent email or message notification, and a reminder is generated in the system;

[0123] High risk corresponds to high alarm level, and the first notification mechanism is instant messaging tool notification, and the responsible person is directly notified;

[0124] The alarm information includes project name, risk type, risk level, specific problem description, etc.

[0125] The notification content includes the key content of the alarm information, such as project name, risk type, risk level, etc. According to the alarm level and risk content, different notification rules are configured, and the responsible person receives the notification and takes corresponding measures according to the disposal suggestion;

[0126] Specifically, according to the disposal method steps and methods, the disposal link is automatically followed up, whether each link is disposed of according to the production project management specification is automatically checked, and the disposal link is displayed in a visual way. Intuitive and efficient support for production project risk control.

[0127] It should be noted that the intelligent alarm system realizes risk grading and accurate notification, quickly responds to exceptions, ensures that the responsible person timely disposes, improves the efficiency and standardization of project risk control, and ensures the smooth progress of the project.

[0128] The above is a schematic scheme of the production project whole-process transparent control index analysis method of the embodiment. It should be noted that the technical scheme of the production project whole-process transparent control index analysis system belongs to the same concept as the technical scheme of the production project whole-process transparent control index analysis method described above. The technical scheme of the production project whole-process transparent control index analysis system in the embodiment is not described in detail. The details can be referred to the description of the technical scheme of the production project whole-process transparent control index analysis method.

[0129] The production project whole-process transparent control index analysis system in the embodiment includes:

[0130] The data integration module is configured to obtain multi-source data, and obtain first index data by integrating the multi-source data.

[0131] A collation module is configured to construct a digital production project management model according to the first index data, and collate production project links by using the digital production project management model to obtain a first list of abnormal data;

[0132] A data analysis module is configured to analyze the first index data based on the first list of abnormal data by using a first natural language processing method combined with a first rule to obtain second abnormal data;

[0133] An alarm information generation module is configured to generate alarm information according to the second abnormal data, and trigger a first notification mechanism to notify relevant personnel to take control measures.

[0134] The embodiment also provides a computer device suitable for production project whole-process transparent control index analysis, which comprises:

[0135] A memory and a processor, wherein the memory is configured to store computer executable instructions, and the processor is configured to execute the computer executable instructions to implement the production project whole-process transparent control index analysis method according to the above embodiment.

[0136] The embodiment also provides a storage medium having a computer program stored thereon, and the program is executed by a processor to implement the production project whole-process transparent control index analysis method according to the above embodiment.

[0137] The storage medium according to the embodiment and the production project whole-process transparent control index analysis method according to the above embodiment belong to the same inventive concept, and the technical details not described in the embodiment can be referred to the above embodiment, and the embodiment has the same beneficial effects as the above embodiment.

[0138] From the above description about the embodiments, those skilled in the art can clearly understand that the present application can be realized by means of software and necessary universal hardware, and of course can also be realized by hardware. Based on such understanding, the technical solutions of the present application can be embodied in the form of a software product, and the computer software product can be stored in a computer readable storage medium, such as a floppy disk, a ROM, a RAM, a FLASH, a hard disk or an optical disk, and includes a plurality of instructions to make a computer device (which can be a personal computer, a server or a network device, etc.) execute the methods of various embodiments of the present application.

[0139] It should be noted that the above examples are only used to illustrate the technical solutions of the present application but not to limit the present application. Although the present application is described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present application can be modified or equivalently replaced, without departing from the spirit and scope of the technical solutions of the present application, and all of them should be covered in the scope of the claims of the present application.

Claims

1. A method for analyzing indicators of transparent control over the entire process of a production project, characterized in that: include: Acquire multi-source data, and obtain first indicator data by integrating the multi-source data; Building a digital production project management model based on the first indicator data, and using the digital production project management model to check the production project links to obtain a first abnormal data list; Based on the first abnormal data list, analyzing the first indicator data using a first natural language processing method combined with a first rule to obtain second abnormal data; Generate an alarm message based on the second abnormal data, and trigger the first notification mechanism to notify relevant personnel to take control measures.

2. A method for analyzing indicators of transparent control over the entire production process of a production project according to claim 1, characterized in that: Integrating the multi-source data includes: Using the first matching method to match multi-source data to obtain key information; The indicators in the key information are sorted and assigned values ​​to obtain the first indicator data for storage.

3. A method for analyzing indicators of transparent control over the entire production process of a production project according to claim 2, characterized in that: Verification of production project links includes: According to the preset rules, each entity and corresponding relationship in the digital knowledge graph model is checked; If the entity or the corresponding relationship does not conform to the preset rule, the first abnormal data is recorded and a first abnormal data list is generated.

4. A method for analyzing indicators of transparent control over the entire production process of a production project according to claim 3, characterized in that: Analyzing the first indicator data includes: Scanning and capturing the first indicator data using a first natural language processing method; The keywords are used for matching, and the first indicator data is transformed and analyzed according to the first rule to obtain the second abnormal data.

5. A method for analyzing indicators of transparent control over the entire production process of a production project according to claim 4, characterized in that: Generating alarm information according to the second abnormal data includes: Define risk levels, warning levels, and the corresponding relationships between risk levels and warning levels; Perform risk assessment on the second abnormal data according to the risk level to obtain a risk assessment result; According to the risk assessment results, the corresponding level of alarm is triggered and the alarm information is obtained.

6. A method for analyzing indicators of transparent control over the entire production process of a production project according to claim 1 or 2, characterized in that: Matching multi-source data using the first matching method includes: Performing string matching and feature code verification on the multi-source data to obtain first multi-source data; Preprocessing the first multi-source data to obtain key information; Build a keyword library for key information of analytical indicators, match keywords, and extract entity information elements; Use word embedding to convert the text into a vector representation, train the semantic matching model, and obtain the first semantic matching model; Calculating the similarity between the searched text to be spliced ​​and the entity information element using the first semantic matching model; Determine the similarity between two entities; If the similarity is greater than the first threshold, the entities are identical, and the corresponding list records are concatenated.

7. A method for analyzing indicators of transparent control over the entire production process of a production project according to claim 2, characterized in that: Also includes: Convert multi-source data into a unified feature vector; Use keywords as nodes to construct associated subgraphs, and map the subgraphs to the relationship nodes in the relationship network to achieve the association of multi-source data; For nodes with similar semantics, the shortest semantic distance is calculated to fuse them and obtain the association relationship between multi-source data.

8. A production project full process transparent control index analysis system, applying a production project full process transparent control index analysis method according to any one of claims 1 to 7, characterized in that: include: A data integration module is used to obtain multi-source data and obtain first indicator data by integrating the multi-source data; a verification module, configured to construct a digital production project management model based on the first indicator data, and verify the production project links using the digital production project management model to obtain a first abnormal data list; a data analysis module, configured to analyze the first indicator data based on the first abnormal data list using a first natural language processing method in combination with a first rule to obtain second abnormal data; An alarm information generation module is used to generate alarm information according to the second abnormal data and trigger the first notification mechanism to notify relevant personnel to take control measures.

9. A computer device, characterized in that: include: A memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of a method for analyzing indicators of transparent control over the entire process of a production project as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that A computer program is stored thereon, and when the computer program is executed by a processor, the steps of a method for analyzing indicators of transparent control over the entire process of a production project as described in any one of claims 1 to 7 are implemented.

Images