VVA method for calculating and analyzing class model

Abstract

The invention discloses a method for calculating and analyzing VVamp of a class model. According to the method A, a problem analysis module, a conceptual model design module, a mathematical logic model design module and a software program model design module are included, the problem analysis module is used for analyzing application requirements of the model and carding practical problems solved in application and expected effects of the model, and the invention relates to the technical field of computational analysis models. Calculating the VVamp of the analysis class model; the method A comprises the following steps: VVamp; the verification, confirmation and authentication of the A evaluation module can comprehensively evaluate the logic and mathematical accuracy of the model and the performance of the model in an expected application range, thereby improving the quality and credibility of a model output result.

Description

Technical Field

[0001] The present invention relates to the technical field of computational analysis models, and in particular to a VV&A method for computational analysis models. Background Art

[0002] VV&A mainly solves the problem of model credibility evaluation. There is no completely consistent systematic expression of the concept of credibility at home and abroad. Drawing on the definition of simulation system credibility, there are roughly three types of definitions from an engineering perspective. Credibility definition based on similarity: Similarity describes the degree of approximation between systems. It is a function of the number of similar units between systems, the values ​​of similar elements, and the total coefficient of influence of each similar element on system similarity. Simulation system credibility refers to the degree of credibility of the simulation system as a similar alternative system to the target system, which can reproduce the target system at the overall structure and behavior level under specific modeling and simulation purposes and meanings. Credibility definition based on confidence: Confidence defines people's confidence in accepting a certain simulation result from a probability theory perspective, and describes simulation credibility with confidence intervals, confidence levels, and confidence levels. Credibility definition based on errors: A more common view is that simulation credibility depends on the accuracy of simulation results, including the degree of refinement and accuracy of result representation. Error analysis methods are usually used to study this type of simulation credibility. Simulation credibility refers to the degree of trust that users of the simulation system have in the correctness of the results of the simulation experiment / test in solving a specific problem. Combining the above three types of credibility definitions, model credibility can be described as: the degree to which the model correctly reflects the real world in terms of process, phenomena and results relative to a specific application purpose. However, the accuracy and credibility of the computational analysis models currently used are relatively poor, and there is an urgent need for accurate and reliable computational analysis models. Summary of the invention

[0003] To achieve the above objectives, the present invention is implemented through the following technical solutions: a VV&A method for computational analysis models, including a problem analysis module, a conceptual model design module, a mathematical logic model design module, and a software program model design module.

[0004] Preferably, the problem analysis module is used to analyze the application requirements of the model, and to sort out the actual problems it solves in the application and the expected role it plays.

[0005] Preferably, the conceptual model is designed to describe the problem, form a universally recognized and unambiguous conceptual model, and define the application problem and the usage boundaries of its concepts.

[0006] Preferably, the mathematical logic model is designed to perform logical abstraction on concepts based on mathematical and physical principles, forming a formalized logical description of a computational analysis model, and describing the core idea and mathematical logic of the concept.

[0007] Preferably, the software program model is designed to describe the mathematical logic of the model using a standardized computer program language to form a computer code or runtime library that is executable by a computer, has input and output interfaces, and can be integrated and deployed.

[0008] The present invention provides a VV&A method for computing and analyzing models. It has the following beneficial effects: The VV&A method of computational analysis models can comprehensively evaluate the model's logic, mathematical accuracy, and performance within the intended application range through verification, confirmation, and certification of the VV&A evaluation module, thereby improving the quality and credibility of the model's output results. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 It is a schematic diagram of the VV&A process framework structure based on the "V-in-V" model of the present invention. DETAILED DESCRIPTION

[0010] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0011] For the first embodiment, please refer to Figure 1 The present invention provides a technical solution: a VV&A method for computing and analyzing models. The VV&A of computing and analyzing models runs through the entire process of demand analysis, design, development and application of model construction, that is, in problem analysis, conceptual model design, mathematical logic model design, and software program model design, the corresponding analysis or design results need to be verified and validated. Therefore, the verification link of the "V" model of the architecture design can be moved forward to form a "V-in-V" model, such as Figure 1 As shown in the figure, on the left is the VV&A verification of the results of each link of model building, and on the right is the integration of the final results of the model into the actual application activities on this basis, and the joint tests are gradually confirmed from the bottom up: ① Carry out internal joint tests to verify the realization of the functional performance described in the conceptual model and the mathematical logic model, as well as the correctness of the interconnection and interoperability interfaces with other related systems and models; ② Carry out field joint tests, combine typical scenarios and application activities, and systematically verify the computing support capabilities of the model; ③ Carry out application testing, combine the actual application of real data, and verify the degree to which the model meets the support requirements of application activities; ④ Carry out comprehensive confirmation, comprehensively evaluate the credibility of the model based on the test results at each level, and confirm its usability.

[0012] VV&A steps for computational analysis models: 1. Main steps of model VV&A According to the "V" model, the computational analysis model VV&A is divided into 5 steps.

[0013] (1) VV&A of demand analysis: Targeting the problem analysis phase. VV&A is to conduct in-depth analysis and decomposition of application requirements based on current technology levels, laws and regulations, and historical experience and knowledge, analyze and verify the possibility of solving actual problems and the possibility of expected effects, and confirm the needs of the model.

[0014] (2) V&V of conceptual model: V&V is to evaluate and calibrate the accuracy of the conceptual model, the rationality of the scope of application, and the unambiguity of the model representation based on the results of demand analysis in a specific application scenario.

[0015] (3) V&V of logical model: V&V is aimed at the design of mathematical logical model. V&V is to evaluate and verify the relationship between mathematical logical model and conceptual model, the accuracy of modeling method, and the consistency of input and output.

[0016] (4) V&V of program model: V&V is to check and verify the code specifications, functional performance indicators, interaction protocols and interface specifications, model output rationality, and integrability of the developed model.

[0017] (5) VV&A of integrated applications: For the integrated application verification phase. VV&A is to integrate the model and deploy it into the actual application verification environment, and gradually complete the model interaction performance verification, technical implementation correctness verification, functional rationality and completeness verification, model requirement satisfaction evaluation, acceptability evaluation and confirmation, etc.

[0018] 2. Model V&V Method Technology Model V&V technology can be divided into four categories: non-standard V&V technology, standard V&V technology, static V&V technology, and dynamic V&V technology.

[0019] (1) Non-standard V&V technology: relies on manual reasoning and subjective judgment, without strict mathematical description and analytical reasoning.

[0020] (2) Standardized V&V technology: technology based on mathematical reasoning and calculation to verify the correctness of the constructed model, usually used as a supplement to non-standardized V&V technology.

[0021] (3) Static V&V technology: used to evaluate the correctness of static model design and source code. It does not require running the model and can reveal information such as model structure, model data flow, control flow, and syntax.

[0022] (4) Dynamic V&V technology: The model needs to be run and the credibility of the model is assessed based on its performance.

[0023] For VV&A of requirement analysis, non-standard V&V techniques can be used, such as expert consultation and meeting review to evaluate the requirement analysis documents.

[0024] For the V&V of the conceptual model, non-standard V&V technology can be used to verify the conceptual model through expert consultation, meeting review, questionnaire, item-by-item scoring, etc. The research field of the experts should be reasonable.

[0025] For the V&V of mathematical logic models, a combination of non-standard V&V technology and standard V&V technology can be used to verify the rationality of the overall logic of the model and the correctness of each mathematical logic module within the model through expert consultation, meeting review, questionnaires, etc. The research field of experts should be reasonable.

[0026] For the V&V of software program models, a combination of non-normative V&V technology and static V&V technology can be used to carry out the V&V. Software tools can be used to verify the standardization of the code, and the rationality of the model interface can be verified through manual code inspection and standard compliance verification. The model's demand response, mathematical logic, and integrability can be verified through expert consultation and meeting review.

[0027] For VV&A of integrated applications, a combination of non-standard V&V technology and dynamic V&V technology can be used to carry out the test. The model can be integrated into the actual application verification environment, specific test scenarios can be designed, and the model can be run. The model's operating efficiency, functional completeness, and degree of satisfaction of requirements can be verified through operation monitoring and data analysis. The model's accuracy and reusability can be analyzed through expert review to confirm the model's credibility.

[0028] Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field and related fields without creative work should fall within the scope of protection of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention are implemented according to the conventional means in the field unless otherwise specified and limited.

Claims

1. A VV&A method for a computational analysis model, characterized in that: It includes problem analysis module, conceptual model design module, mathematical logic model design module and software program model design module.

2. A VV&A method for computing and analyzing models according to claim 1, characterized in that: The problem analysis module is used to analyze the application requirements of the model, sort out the actual problems it solves in the application and the expected role it plays.

3. The VV&A method for computing and analyzing models according to claim 1, characterized in that: The conceptual model is designed to describe the problem, form a universally recognized and unambiguous conceptual model, and define the application problem and the usage boundaries of its concepts.

4. The VV&A method for computing and analyzing models according to claim 1, characterized in that: The mathematical logic model is designed to perform logical abstraction on concepts based on mathematical and physical principles, forming a formalized logical description of computational analysis models, and describing the core ideas and mathematical logic of the concepts.

5. The VV&A method for computing and analyzing models according to claim 1, characterized in that: The software program model is designed to describe the mathematical logic of the model using a standardized computer program language to form a computer code or runtime library that is executable by a computer, has input and output interfaces, and can be integrated and deployed.

Images