Water conservancy project safety management and control model library system and method based on user credibility

Abstract

The present invention relates to a water conservancy project safety management and control model library system and method based on user credibility, wherein the model library system comprises a safety management and control model service unit, a safety management and control model reference unit and a model user credibility calculation unit. The safety management and control model service unit is used to provide storage, management and call of safety risk prediction and hidden danger warning models in the model library; the safety management and control model reference unit is used to memorize and share the existing safety management and control model call results in the system, and can reproduce the calculation process of the risk prediction or hidden danger warning model; the model user credibility calculation unit is used to construct a professional character portrait of the user of the model, and evaluate the credibility of the user model calculation results from a professional perspective. The present invention not only realizes the management and use of various models in the water conservancy project safety management and control decision-making process, but also assigns a credibility coefficient to the model user, thereby improving the professional credibility of the water conservancy project safety management and control decision-making process.

Description

Technical Field

[0001] The present invention relates to a water conservancy project safety risk management and control model library system and method, and in particular to a water conservancy project safety risk management and control model library system and method based on user credibility. Background Art

[0002] With socioeconomic development, water conservancy projects such as reservoirs, dams, and hydropower stations are being built at a rapid pace. The overall construction technology of water conservancy projects is relatively complex, susceptible to various factors such as hydrogeology and topography. These projects are characterized by high capital investment and high construction difficulty, requiring effective coordination among multiple departments to ensure high-quality completion of various construction tasks. Furthermore, every aspect of water conservancy projects carries immeasurable potential risks, which complicates project safety management.

[0003] Due to the difficulties in safety management and control of water conservancy projects, we can reduce the difficulty of project safety management and control by conducting risk prediction, hidden danger warning and other auxiliary decision-making work in advance for the project construction process, and proposing scientific and effective supervision measures. However, risk prediction and hidden danger warning in project construction rely on the knowledge reserves and project experience of relevant experts. Each expert's evaluation information has different focuses. Based on the materials submitted by the user, the credibility coefficient is calculated after review and approval. Finally, the risk prediction, hidden danger warning and other results of all users are calculated, which can reduce the influence of subjective factors on objective results and obtain more scientific decision-making results. Summary of the Invention

[0004] Purpose of the invention: The first purpose of the present invention is to provide a security control model library system in which model call information can be memorized and shared, and the credibility of model calculations such as risk prediction and hidden danger warning can be improved; the second purpose of the present invention is to provide a security control model library method based on user credibility.

[0005] Technical solution: The water conservancy project safety management and control model library system based on user credibility of the present invention includes a safety management and control model service unit, a safety management and control model reference unit and a model user credibility calculation unit;

[0006] The security control model service unit is used to store, manage, and call various security risk prediction, hidden danger warning, and other models in the model library;

[0007] The safety management and control model reference unit is used to memorize and share the parameters, status, and results of the model during operation, and to reproduce the entire process of model calculation and analysis, such as risk prediction and hidden danger warning.

[0008] The model user credibility calculation unit is used to create a professional profile of the model user and evaluate the credibility of user evaluation information from a professional perspective.

[0009] Furthermore, the security management model service unit has built a model library system architecture based on risk prediction, hidden danger warning and other services, providing model users with model calculation and analysis functions, and providing model administrators with model service management functions, which can add, query, delete and other operations on the model.

[0010] The safety management and control model reference unit includes the model work process attribute extraction, storage and model work process reference functions; the model work process is the operating parameters and process attributes of the risk prediction, hidden danger warning and other models already existing in the model library system, and the parameters and attributes involved in the entire process from the beginning to the end of the model work.

[0011] The process attributes before the model work starts include semi-structured data attributes such as model user information, number of model variables and sample data; the process attributes after the model work starts include semi-structured data attributes such as parameter test values, model coefficients and model usage date; the result attributes after the model work ends include structured data such as actual output variable results of the model and model satisfaction.

[0012] The model's user credibility calculation unit includes functions such as user data collection, administrator review and certification, and professional profile construction. The user professional profile data categories include personal information and project experience. Personal information includes structured data attributes such as education, professional qualifications, and academic authority. Project experience includes semi-structured data attributes such as length of professional experience, relevant project experience, and project leadership.

[0013] The model administrator reviews and evaluates all user profiles, generates user tags based on the audit and certification results, and builds user profiles. The administrator also assigns weighting factors to each attribute based on the actual model, and comprehensively calculates the user's credibility coefficient.

[0014] To summarize, the safety management and control model service unit builds a model library system architecture based on risk prediction, hidden danger warning and other services, and provides model library management functions; the safety management and control model reference unit is mainly responsible for accessing the attributes of the completed model work process, providing users with parameters, status and results in the model work process, and realizing the model work process reference function; the model user credibility calculation unit is mainly responsible for collecting and reviewing model user information, generating user tags and building user portraits, introducing information credibility coefficients for users, and improving the scientific nature of the results.

[0015] The method for using the water conservancy project safety management and control model library based on user credibility includes the following steps:

[0016] (1) The safety management model service unit builds a model library system architecture based on risk prediction, hidden danger warning and other services, and provides model library management functions;

[0017] (2) Before the user calls the model for the first time, the model user credibility calculation unit will be automatically activated, and the user will submit relevant certificates and explain project experience. The user will be reviewed and a user tag will be generated for portrait authentication, and the user's professional credibility coefficient will be comprehensively calculated;

[0018] (3) When the user calls the model, he can wake up the security control model reference unit to store and extract the model working process attributes, remember and share the parameters, status and results of the model working process, and realize the model working process reference function;

[0019] (4) The user model call is completed and the actual output of the model is obtained.

[0020] Furthermore, in step (2), after the model user credibility calculation unit is automatically awakened, the model user submits relevant certificates and self-explanations according to the designed UI interface prompts. The personal information user portrait attribute is authenticated by submitting a degree certificate and a professional qualification certificate, respectively, and the academic authority certification can be verified by submitting academic research results such as the number of published papers and the number of cited papers; the project experience user portrait attribute is reviewed and certified by the model administrator by selecting the time spent in the profession and self-explanation of project experience; finally, the administrator of each model sets an appropriate weight factor β for each user portrait attribute according to the actual situation of the project. i , where ∑β i = 1. If the review score of the certification materials submitted by the model user is ω i (ω i ≤1), in order to comprehensively consider all model users, the credibility coefficient of the user is calculated according to formula (1) as α i , the user's credibility coefficient can be shared among different model libraries of the project.

[0021]

[0022] After obtaining the user credibility coefficient, the risk hidden danger information to be predicted by the model is comprehensively calculated according to formula (2), where n is the number of users of the model (i.e., the number of experts), α i is the credibility coefficient of model user i, D i is the result of model user i's treatment of predicted risk hazards, and D is the evaluation result of the risk hazards to be predicted.

[0023]

[0024] Furthermore, in step (3), the user can freely choose whether to wake up the security control model reference unit. The model work process attribute storage adopts a hybrid storage method of relational database management system and XML data document. Using structured data such as relational database model dictionary, the model work process attribute document (dynamic process of model operation) can be stored using XML data document. If wake up is selected, the work process attribute of the model service can be extracted through the XML parser to realize the model work process reference function, and finally presented to the user using the UI component.

[0025] Beneficial effects: Compared with the existing model library technology, the present invention has the following significant advantages: by combining the two hybrid data storage methods of relational database and XML data document, dynamic access to model work process attributes is realized, and the current problem of model work information sharing such as risk prediction and hidden danger warning is solved; another advantage is that it takes into account the user's emphasis on evaluation information, gives the user a credibility coefficient, and comprehensively calculates the evaluation information, which reduces the influence of subjectivity on the results and improves the scientific nature of the results. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is the system function structure diagram of the present invention

[0027] Figure 2 This is the system function flow chart of the present invention

[0028] Figure 3 This is the overall system architecture diagram of the present invention DETAILED DESCRIPTION

[0029] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0030] The core of the present invention is a water conservancy project safety management and control model library system and method based on user credibility, which not only realizes the sharing and reference of model working process information, but also improves the professional credibility of model calculations such as water conservancy project safety risk prediction and hidden danger warning to a certain extent. Figure 1 As shown, the water conservancy project safety management and control model library system based on user credibility of the present invention includes a safety management and control model service unit, a safety management and control model reference unit and a model user credibility calculation unit.

[0031] The safety management and control model reference unit focuses on a series of work process attributes included in model calculations such as risk prediction and hidden danger warning. The process attributes before and after the start of the model work and before and after the end of the model work are stored in a mixed manner through structured data attributes and semi-structured data attributes, so as to realize the memory and sharing of the model work process and calculation results, and reproduce the entire process of model calculation and analysis such as risk prediction and hidden danger warning.

[0032] The model user credibility calculation unit takes into account the different emphases of users on evaluation information. Through the personal information and project experience information submitted by users, the model administrator sets different weight factors for different attributes and comprehensively calculates the user's credibility coefficient, reducing the impact of subjective factors on the evaluation results and improving the scientific nature of the results.

[0033] Figure 2 It is a specific implementation flow chart of the present invention, which includes the following steps:

[0034] Step 1: The security management model service unit builds a model library system architecture based on decision-making services such as risk prediction and hidden danger warning. It not only provides model users with model calculation and analysis functions, but also provides model administrators with model data storage, management and call functions, and can add, query, delete and other operations on the model.

[0035] like Figure 3 As shown, the present invention uses a microservices architecture to deconstruct system functions into services such as model management, model service encapsulation and invocation, model work process reference, and model user credibility calculation. Each service is deployed relatively independently, and the operation of a microservice does not affect other microservices in the system. The present invention adopts a message-driven microservices architecture, in which each microservice completes message subscription in the message service middleware. RabbitMQ is used as the message service middleware to implement the subscription-publish model of each service.

[0036] The client application is based on the Vue framework and uses ElementUI as its UI framework. Requests for retrieving models are specified as GET, while requests for adding and deleting models are specified as POST. Data transmission is encapsulated using the JSON class, and Axios is used to send Ajax requests for front-end and back-end communication. The Django view class receives the parameters requested by the front-end, calls the corresponding model management microservice via RPC, uses the ORM to operate the database, and finally returns the results to the front-end using the JSONResponse method.

[0037] Step 2: Before the user calls the model for the first time, the model user credibility calculation unit is automatically awakened, the credibility coefficient is introduced, relevant proof and project experience are submitted, and the model administrator reviews the materials, generates user tags, builds a professional portrait, and comprehensively calculates the user's professional credibility.

[0038] This unit is developed based on the web front-end framework Vue. HTML is responsible for web page elements, CSS is responsible for rendering pages, and JavaScript is responsible for data processing and interaction. ElementUI is used as the UI framework, which can improve web development efficiency to a certain extent. The personal information user portrait attribute is verified by submitting a degree certificate and a professional qualification certificate to verify their educational qualifications and professional qualifications respectively. Academic authority certification can be verified by submitting academic research results such as the number of published papers and the number of citations. The project experience user portrait attribute is verified by selecting a category and self-explanation of relevant project experience in the selected drop-down box in the UI, and then reviewed and certified by the model administrator.

[0039] The model user credibility calculation unit of the present invention has a model administrator identity. When the model is released and used, the administrator of each model sets an appropriate weight factor β for each user portrait attribute according to the actual situation of the project. i , where ∑β i = 1. If the review score of the certification materials submitted by the model user is ω i (ω i ≤1), in order to comprehensively consider all model users, the credibility coefficient of the user is calculated according to formula (3) as α i , the user's credibility coefficient can be shared among different model libraries of the project.

[0040]

[0041] After obtaining the user credibility coefficient, the risk hidden danger information to be predicted by the model is calculated according to formula (4), where n is the number of users of the model (i.e., the number of experts), α i is the credibility coefficient of model user i, D i is the result of model user i's treatment of predicted risk hazards, and D is the evaluation result of the risk hazards to be predicted.

[0042]

[0043] This embodiment takes the credibility coefficient in the improved LEC method in the management and control of a water conservancy project as an example. According to the actual situation of the project, a suitable weight factor has been set for each expert user portrait attribute, as shown in Table 1 below.

[0044] Table 1 Table of weight factors of user portrait attributes in a project management model

[0045]

[0046] We selected three experts from the project as an example, and conducted certification review on the materials submitted by each expert according to the criteria in Table 1. The final material certification results are shown in Table 2 below.

[0047] Table 2 Scores of profile attribute materials of users in the control model

[0048] Do you have experience in similar projects? Professional Qualification Length of professional experience Scoreω Expert No. 1 Yes (+0.2) Engineer (+0.2) Six years (+0.2) 0.6 Expert No. 2 Yes (+0.2) Senior Engineer (+0.4) Eight years (+0.3) 0.9 Expert No. 3 None (+0) Engineer (+0.2) Five years (+0.2) 0.4

[0049] In Table 2, we can clearly see the specific material certification situation of each expert. The credibility coefficient of each expert can be calculated according to formula (1), and the expert credibility vector w = [0.316, 0.474, 0.211] is obtained.

[0050] Step 3: When calling the model, the user can freely choose whether to wake up the security control model reference unit. The model working process attribute storage adopts a hybrid storage method of relational database management system and XML data documents, which remembers and shares the parameters, status and results of the model working process, realizing the model working process reference function;

[0051] The model dictionary stores the basic attributes of the model itself and is relational data that can be described in a structured manner. Therefore, it can be stored directly in a relational database management system. The model operation property document describes the work process attributes involved in the model calculation from the beginning to the end in an XML data document. The structure of the model operation property document varies from model to model. This document is ultimately stored in a two-dimensional table field of the text large field type in the relational database management system and can be accessed through an XML parser. In the implementation of this model library, Web Services and widely used common protocols (such as HTTP and SOAP) and XML format are used for service encapsulation and data exchange for risk prediction, hidden danger warning and other models.

[0052] Step 4: The model library system provides two model calling methods: Web interface and API, which can be called in the browser or external application. By passing in the various parameters required by the model, model analysis and model calculation are realized, and finally risk prediction, hidden danger warning and other results are returned.

Claims

1. A water conservancy project safety management and control model library system based on user credibility, characterized in that: It includes a security control model service unit, a security control model reference unit, and a model user credibility calculation unit; The security control model service unit is used to provide storage, management and call of security risk prediction and hidden danger warning models in the model library; The safety management and control model reference unit is used to memorize and share the parameters, status and results of each model working process, and can reproduce the entire process of risk prediction and hidden danger warning model calculation and analysis; The model user credibility calculation unit is used to construct a user professional profile for the model user and evaluate the credibility of the user model calculation results from a professional perspective.

2. The water conservancy project safety management and control model library system based on user credibility according to claim 1 is characterized by: The safety management and control model service unit has built a model library system architecture for water conservancy project safety management, providing model users with model calculation and analysis functions, and providing model library administrators with model management functions, which can add, query, and delete models.

3. The water conservancy project safety management and control model library system based on user credibility according to claim 1 is characterized by: The security control model reference unit includes the model work process attribute extraction, storage and model work process reference functions; The model working process includes the operating parameters and process attributes of the risk prediction and hidden danger warning models already in the model library system, and the parameters and attributes involved in the entire process from the beginning to the end of the model working; The process attributes before the model work starts include semi-structured data attributes such as model users, number of model variables and sample data; The process attributes after the model work starts include semi-structured data attributes such as parameter test values, model coefficients and model use date; The result attributes after the model work is completed include structured data such as the actual output variable results of the model and the model satisfaction.

4. The water conservancy project safety management and control model library system based on user credibility according to claim 1 is characterized by: The model user credibility calculation unit includes user data collection, administrator review and authentication, and professional portrait construction functions; The user data collection includes personal information and project experience data types, where personal information includes structured data attributes such as education, professional qualifications, and academic authority; while project experience includes semi-structured data attributes such as length of professional experience, experience with similar projects, and experience in leading projects; The administrator review and certification is to review and evaluate all relevant information collected. At the same time, the administrator assigns a weight factor to each attribute based on the actual model to facilitate the subsequent calculation of the credibility coefficient; The data for constructing the professional portrait comes from the self-submission of the model user, and then the user label is generated based on the administrator's review results, and finally the user's credibility coefficient is obtained by comprehensive calculation.

5. The method for using the water conservancy project safety management and control model library system based on user credibility according to claim 1 is characterized in that: The following steps are involved: (1) The safety management and control model service unit builds a model library system architecture based on risk prediction and hidden danger warning model services, and provides model library management functions; (2) Before the user calls the model for the first time, the model user credibility calculation unit will be automatically activated, and the user will submit relevant proof and explain project experience. The user will review the information, generate user tags, build a user profile, and comprehensively calculate the user's professional credibility coefficient; (3) When the user calls the model, he can wake up the security control model reference unit to realize the storage and extraction of the model working process attributes, remember and share the parameters, status and results of the model working process, and realize the model working process reference function; (4) The user model call is completed and the actual output of the model is obtained.

6. The method for using the water conservancy project safety management and control model library system based on user credibility according to claim 5 is characterized in that: In step (2), after automatically waking up the model user credibility calculation unit: The personal information user profile attribute is verified by submitting relevant qualification certificates to verify their educational qualifications and professional qualifications. The academic authority is evaluated by the administrator through comprehensive assessment of the number of published papers and citations. The project experience user profile attribute is verified by the administrator through the professional time and self-explanation of relevant project experience. The administrator will review and authenticate the collected materials, generate user tags and build user portraits, and assign attribute weights based on the actual project situation, setting appropriate weight factors β for each user portrait attribute. i , where ∑β i =1, if the review score of the certification materials submitted by the model user is ω i , where ω i ≤1, then the credibility coefficient of the model user is calculated according to formula (1) as α i , the user's credibility coefficient can be shared among different model libraries of the project, After obtaining the user credibility coefficient, the risk hidden danger information to be predicted by the model is calculated according to formula (2). Where n is the number of users of the model, i.e. the number of experts, α i is the credibility coefficient of model user i, D i is the result of model user i's treatment of predicted risk hazards, and D is the final result of the risk hazards to be predicted.

7. The method for using the water conservancy project safety management and control model library based on user credibility according to claim 5 is characterized by: In step (3), the user can freely choose whether to wake up the security control model reference unit. The model working process attribute storage adopts a hybrid storage method of relational database management system and XML data document. If wake-up is chosen, the model working process reference function will be realized by extracting the model working process attributes, which can be extracted by XML parser and finally presented on the UI interface.

Images