Confidence assessment model construction method, confidence assessment method and device

Abstract

The present invention provides a confidence assessment model construction method, a confidence assessment method and a device, which relate to the technical field of confidence assessment, including: obtaining a preset element set, element relationship and confidence assessment index set in a test space; using a preset formalized assessment model and element relationship, performing interactive relationship analysis and processing on the element set, determining the interactive relationship between each element and the test variable, and determining the test space confidence symbol representation system using the interactive relationship; using the test space confidence symbol representation system, respectively performing indicator analysis and processing on the confidence assessment indicators of each dimension in the confidence assessment index set, determining the indicator formal representation model corresponding to each confidence assessment indicator, and integrating the indicator formal representation model to construct a confidence assessment model. The present invention can obtain a high-dimensional and definite credibility assessment result of the test space.

Description

Technical Field

[0001] The present invention relates to the technical field of confidence assessment, and in particular to a confidence assessment model construction method, a confidence assessment method and a confidence assessment device. Background Art

[0002] In the current engineering practice of equipment development, it is necessary to evaluate the credibility of multi-source information and models in various scenarios, and use confidence representation and measurement as the basis for judgment to determine whether the test space construction is accurate, whether the functions meet the needs, whether the correctness of the conclusions can meet the needs of decision-making and analysis, and whether the models at different development stages of the prototype system accurately reflect the design features. These are important issues in the equipment development process.

[0003] At present, relevant technologies have proposed that through traditional document-based system design, natural language-based documents such as user needs and design plans can be used as information carriers for system engineering cognition construction to organize and manage the planning, research, design, manufacturing, testing and use of complex systems. However, the above scheme is time-consuming, labor-intensive and error-prone in information integration. If model-based systems engineering (MBSE) is adopted, digital modeling is used instead of writing documents to design system plans, and all nouns, verbs, adjectives, parameters, etc. in the design documents that describe the system structure, function, performance, and specification requirements are converted into digital models for expression, when imperfect cognition is solidified in the model code, it will have a negative impact on the entire life cycle of system engineering through the model library. In addition, the standardization and normalization of system modeling and verification are insufficient, there is a lack of a modeling technology system covering all engineering disciplines, an integrated system design and verification platform has not been formed, and there is a lack of systematic application based on actual engineering models, which leads to low accuracy of the evaluation results. Summary of the invention

[0004] In view of this, an object of the present invention is to provide a confidence assessment model construction method, a confidence assessment method and a device, which can significantly improve the assessment accuracy.

[0005] In the first aspect, an embodiment of the present invention provides a method for constructing a confidence assessment model, the method comprising: obtaining a preset element set, element relationship and confidence assessment indicator set in an experimental space, wherein the element set comprises: a meta set, a confidence set, an algorithm set and a threshold set, and the element relationship comprises: equivalence, inclusion and true inclusion; through a preset formalized evaluation model, utilizing the element relationship, performing interaction relationship analysis and processing on the element set, determining the interaction relationship between each element and the experimental variable, and determining the test space confidence symbolic representation system using the interaction relationship, wherein the experimental variable is a variable that can be directly measured in the experimental space; utilizing the test space confidence symbolic representation system, performing indicator analysis and processing on the confidence assessment indicators of each dimension in the confidence assessment indicator set, determining the indicator formal representation model corresponding to each confidence assessment indicator, and integrating the indicator formal representation model to construct a confidence assessment model.

[0006] In one embodiment, a confidence assessment indicator set includes: multi-source information confidence, model confidence and prediction result confidence. The test space confidence symbol representation system is used to perform indicator analysis and processing on the confidence assessment indicators of each dimension in the confidence assessment indicator set. The step of determining the indicator formal representation model corresponding to each confidence assessment indicator includes: based on the test space confidence symbol representation system, the indicator analysis and processing are performed on the multi-source information confidence, the model confidence and the prediction result confidence, respectively, to determine the multi-source information confidence formal representation model, the model confidence formal representation model and the prediction result confidence formal representation model, wherein the multi-source information confidence formal representation model is used to evaluate the credibility of the simulation data, physical test data, key factors and prior knowledge of the model input, the model confidence formal representation model is used to evaluate the credibility of the model and prototype, and the prediction result confidence formal representation model is used to evaluate the credibility of the model output result.

[0007] In one embodiment, based on the symbolic representation system of experimental space confidence, an indicator analysis and processing is performed on the confidence of multi-source information to determine the steps of a formal representation model of the multi-source information confidence, including: performing indicator analysis and processing on the sample point quality and the consistency of multi-source information through the symbolic representation system of experimental space confidence, and determining the formal representation model of sample point quality and the formal representation model of multi-source information consistency, wherein the sample point quality is a characteristic measure of the sample point, and the consistency of multi-source information is used to represent the consistency of information from different sources relative to the real information when representing the same indicator information; combining the formal representation model of sample point quality and the formal representation model of multi-source information consistency to determine the formal representation model of multi-source information confidence.

[0008] In one embodiment, based on the test space confidence symbolic representation system, an indicator analysis is performed on the model confidence to determine the model confidence formal representation model, including: through the test space confidence symbolic representation system, an indicator analysis is performed on the first model uncertainty and the second model uncertainty to determine the first model uncertainty formal representation model and the second model uncertainty formal representation model, wherein the first model uncertainty is the uncertainty corresponding to the model with parameter changes, and the second model uncertainty is the uncertainty corresponding to the model with unchanged model and input changes; the first model uncertainty formal representation model and the second model uncertainty formal representation model are combined to determine the model confidence formal representation model.

[0009] In one embodiment, based on the test space confidence symbolic representation system, an indicator analysis and processing is performed on the prediction result confidence, and the step of determining the prediction result confidence formal representation model includes: performing indicator analysis and processing on accuracy and 95% confidence through the test space confidence symbolic representation system, and determining the accuracy formal representation model and the 95% confidence formal representation model, wherein the accuracy is used to represent the deviation of the model prediction result and the system prediction result relative to the benchmark data; combining the accuracy formal representation model and the 95% confidence formal representation model to determine the prediction result confidence formal representation model.

[0010] In a second aspect, an embodiment of the present invention further provides a confidence assessment method, the method comprising: obtaining data to be evaluated in a test space to be tested, wherein the data to be evaluated comprises: sample point information, model information, and simulation and test data information; sending the sample point information, model information, and simulation and test data information to a pre-constructed confidence assessment model, respectively, determining multi-source information confidence assessment results, model confidence assessment results, and prediction result confidence assessment results, and determining the multi-source information confidence assessment results, model confidence assessment results, and prediction result confidence assessment results as target confidence assessment results in the test space to be tested; wherein the confidence assessment model is constructed based on any one of the confidence assessment model construction methods provided in the first aspect.

[0011] In the third aspect, an embodiment of the present invention also provides a confidence assessment model construction device, the device comprising: a data acquisition module, which acquires a preset element set, element relationship and confidence assessment indicator set in the test space, wherein the element set comprises: a meta set, a confidence set, an algorithm set and a threshold set, and the element relationship comprises: equivalence, inclusion and true inclusion; an interactive relationship analysis module, which uses the element relationship through a preset formal evaluation model to perform interactive relationship analysis on the element set, determine the interactive relationship between each element and the experimental variable, and use the interactive relationship to determine the test space confidence symbol representation system, wherein the experimental variable is a variable that can be directly measured in the experimental space; a model training module, which uses the test space confidence symbol representation system to perform indicator analysis on the confidence assessment indicators of each dimension in the confidence assessment indicator set, determine the indicator formal representation model corresponding to each confidence assessment indicator, and integrate the indicator formal representation model to construct a confidence assessment model.

[0012] In a fourth aspect, an embodiment of the present invention further provides a confidence assessment device, the device comprising: a data acquisition module, for acquiring data to be evaluated in a test space to be tested, wherein the data to be evaluated comprises: sample point information, model information, and simulation and test data information; a confidence assessment module, for sending the sample point information, model information, and simulation and test data information to a pre-constructed confidence assessment model, respectively, to determine multi-source information confidence assessment results, model confidence assessment results, and prediction result confidence assessment results, and to determine the multi-source information confidence assessment results, model confidence assessment results, and prediction result confidence assessment results as target confidence assessment results in the test space to be tested; wherein the confidence assessment model is constructed based on any one of the confidence assessment model construction methods provided in the first aspect.

[0013] In a fifth aspect, an embodiment of the present invention further provides a server, comprising a processor and a memory, wherein the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement any one of the methods provided in the first aspect.

[0014] In a sixth aspect, an embodiment of the present invention further provides a computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to implement any one of the methods provided in the first aspect.

[0015] The embodiments of the present invention bring the following beneficial effects:

[0016] The embodiment of the present invention provides a confidence assessment model construction method, a confidence assessment method and a device. After obtaining a preset element set, element relationship and confidence assessment index set in the test space, the method uses the element relationship through a preset formalized assessment model to perform interaction relationship analysis and processing on the element set, determine the interaction relationship between each element and the test variable, and use the interaction relationship to determine the test space confidence symbol representation system. Finally, using the test space confidence symbol representation system, the confidence assessment indicators of each dimension in the confidence assessment index set are respectively analyzed and processed to determine the indicator formal representation model corresponding to each confidence assessment indicator, and the indicator formal representation model is integrated to construct the confidence assessment model. The implementation of the present invention can obtain a high-dimensional and definite credibility assessment result of the test space.

[0017] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0018] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0020] Figure 1 A flow chart of a confidence assessment model construction method provided by an embodiment of the present invention;

[0021] Figure 2 A schematic diagram of the structure of a test space confidence symbol representation system provided by an embodiment of the present invention;

[0022] Figure 3 A flow chart of a confidence assessment method provided by an embodiment of the present invention;

[0023] Figure 4 A schematic diagram of the structure of a confidence assessment model building device provided by an embodiment of the present invention;

[0024] Figure 5 A schematic diagram of the structure of a confidence assessment device provided by an embodiment of the present invention;

[0025] Figure 6 A schematic diagram of the structure of a server provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described in combination with the embodiments below. Obviously, the described embodiments are 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.

[0027] At present, systems engineering (SE) is a scientific method for organizing the planning, research, design, manufacturing, testing and use of complex systems. Its core issue is that the engineering group builds collective cognition of complex systems through thinking and practice, and realizes engineering objects in the objective world based on this cognition, that is, the process of understanding and transforming the world; the information carrier for systems engineering to build cognition is a series of natural language-based documents, such as user needs, design plans, etc. Under this model, information integration work is time-consuming, labor-intensive and prone to errors. With the development of modern information technology, compared with traditional document-based system design, Model-Based Systems Engineering (MBSE) uses digital modeling instead of writing documents to design system solutions, and converts all nouns, verbs, adjectives, and parameters in the design documents that describe the system structure, function, performance, and specification requirements into digital model expressions. MBSE uses the authoritative model source and data source of the system, and the models and data that can be continuously transmitted across disciplines and fields during the life cycle to support all activities of the system from concept development to scrapping and disposal. The problems and bottlenecks of MBSE in modeling include: if imperfect cognition is solidified in the model code, it will have a negative impact on the entire life cycle of the system engineering through the model library; the standardization and normalization of system modeling and verification are insufficient; there is a lack of a modeling technology system covering all engineering disciplines; an integrated system design and verification platform has not been formed; there is a lack of systematic application based on actual engineering models, etc.

[0028] Prototype-&Model-Based Systems Engineering (PMBSE, hereinafter referred to as prototype system engineering) is the latest development of system engineering. Its core is to support the digitization of equipment products and the digitization of the development process through digital system models, realize the digital delivery of products, virtual simulation tests of the entire product system, and the digital model of the entire development process. Through multiple rounds of iterations of "product design-virtual testing" with low cost and high efficiency in digital space, the prototypes and models produced during the entire life cycle of development have the ability to integrate, deduce, and verify across stages and virtual reality. At the same time, the prototypes and models need to be integrated into a complex system to support deductive analysis, so as to support the digital development of the entire process of equipment with high quality and efficiency, and achieve innovation, quality improvement, efficiency increase, and cost reduction. The traditional decentralized test and identification methods can no longer meet the current rapid development needs of complex equipment systems. It is necessary to carry out integrated tests from the perspective of the entire system, and realize the optimal configuration of various simulation, test and other resources in the three dimensions of test means, subsystems and equipment life cycle, so as to improve the development efficiency of equipment and shorten the development cycle. Small samples are an important feature of integrated testing: due to the constraints of test costs, development cycles, and safety, it is impossible to conduct large-scale full-system physical tests. Therefore, only a small number of physical tests and a large number of simulation methods can be used to provide key data and verification to support the construction of test space and the rapid development and update and upgrade of complex systems. The test space is a set of all possible states of the entire life cycle of a complex system constructed in a digital space by comprehensively utilizing multi-source information such as test data, simulation models, and engineering knowledge generated at various stages of the life cycle of the complex system. It can also be simplified as the state space of the equipment at various stages. By constructing the test space and applying it to cross-stage and cross-virtual prototype system engineering methods, rapid iteration of complex systems can be achieved, development costs can be reduced, and R&D efficiency can be improved.

[0029] In engineering practice, due to the limitations of test and simulation methods, we often cannot get the real test space, but only the likelihood estimate for the test space. In order to evaluate the credibility of the test space, the comprehensive confidence can be used to characterize and measure it. Traditional confidence is an important concept in statistics. It is used to evaluate the reliability of a statistical inference or estimation, usually expressed in percentage. The comprehensive confidence is a high-dimensional parameter characterization and measurement of the credibility of multi-source information, models and prediction results involved in all possible states of the test space. It contains the information of traditional confidence and expands the evaluation index of the credibility of the test space on this basis. The test space confidence mainly consists of three parts: multi-source information confidence, model confidence and prediction result confidence. Multi-source information confidence is used to evaluate the credibility of information such as simulation data, physical test data, key factors, prior knowledge, etc. input to the model; model confidence is used to evaluate the credibility of models and prototypes, such as the uncertainty of mechanism models and the transparency of machine learning models; prediction result confidence is used to evaluate the credibility of model output results, reflecting the reliability, accuracy and uncertainty of prediction results.

[0030] In the current engineering practice of equipment development, traditional confidence assessment indicators and methods can no longer meet the needs of credibility assessment of multi-source information and models in various scenarios. Whether the test space is accurately constructed, whether the functions meet the needs, whether the correctness of the conclusions can meet the needs of decision-making and analysis, and whether the models of the prototype system at different development stages accurately reflect the design characteristics are all important issues in the equipment development process, which need to be judged by confidence characterization and measurement. The current confidence assessment theory has experienced more than 60 years of development. The academic community has a variety of definitions of confidence assessment. Some believe that it is the degree of similarity between the simulation system and the simulation object; some believe that it is the degree of confidence that the simulation user has in using the simulation system to solve practical problems; and some believe that it is the inverse of the deviation between the simulation data and the reference data. There is also a different understanding of the connotation of confidence in engineering, which leads to inconsistent use of terms. Frequently used similar terms include "credibility", "fidelity", "precision", etc. This chaotic situation has hindered daily academic exchanges and engineering applications, and restricted the further development of confidence assessment theory. Therefore, it is necessary to construct a test space confidence symbol representation system suitable for the evolution of prototype system models. Based on this, the confidence assessment model construction method, confidence assessment method and device provided by the present invention can significantly improve the assessment accuracy by constructing a confidence assessment model of the complex system test space. The system confidence can be verified and confirmed in the early stage of complex system development, which can greatly reduce various risks and reduce the cycle and cost of design changes, thereby solving the problem that the test space cannot be directly obtained, and providing reliability technical support for the test space construction and overall system practice of virtual-reality integrated testing of complex systems.

[0031] See also Figure 1 The flowchart of a confidence assessment model construction method shown in FIG. 1 mainly includes the following steps S102 to S106:

[0032] Step S102, obtaining a preset element set, element relationship and confidence assessment indicator set in the test space, wherein the element set includes: a meta set, a confidence set, an algorithm set and a threshold set, and the element relationship includes: equivalence, inclusion and true inclusion. In one embodiment, this scheme is based on a formalized evaluation model, uses a meta set, a confidence set, an algorithm set and a threshold set as the theoretical basis for confidence assessment, constructs a corresponding confidence assessment model, realizes the symbolic representation of the confidence assessment model, and can apply the symbolic representation to the design of a complex system test space suitable for prototype system engineering; the mechanism model, data and design parameters and task parameters (physical quantities that change when the system is running, such as speed, thrust, etc.) in the physical prototype formed in the design process can be directly encapsulated as "states", and the set consisting of "states" between different systems or different "states" of the same system is called a meta set, wherein the part of the meta set that needs to be confidence assessed is used as a key indicator to provide data support for the subsequent algorithm set calculation process.

[0033] Specifically, the set of simulation modules in the meta-set is called entity meta-set, the set of interaction relationships between simulation modules is called interaction meta-set, the set of attributes of simulation objects (such as inertia, conductivity, shape, color, etc., and in the solid-liquid rocket engine thrust simulation program, the ablation rate belongs to the attributes of the simulation program) is called attribute meta-set; the variables that can be directly measured in the simulation modules and the simulation systems they constitute are called variable meta-sets; in addition, the set that represents the correlation (coupling) between variables is called correlation meta-set; then the meta-set is equal to the union of the above five sets, and these five sets do not intersect with each other.

[0034] In the algorithm set, algorithms are divided into measurement algorithms and comprehensive algorithms. Among them, measurement algorithms are specific methods for attribute calculation, such as deviation calculation, average calculation, variance calculation, etc.; comprehensive algorithms are algorithms for the interaction relationship between systems and the confidence definition relationship, such as information confidence algorithm, physical confidence algorithm, model confidence algorithm, interaction algorithm, etc.

[0035] The threshold set is a set of preliminary qualitative and quantitative evaluations of the model, such as topological shape judgments of model indicators and data integrity judgments of the model; the confidence set is a confidence assessment result that includes multi-source information, models and predicted data; in the entity element, the relationships between elements are: equivalence, inclusion and true inclusion.

[0036] Step S104, through the preset formal evaluation model, using the element relationship, the element set is analyzed and processed for the interaction relationship, the interaction relationship between each element and the test variable is determined, and the test space confidence symbol representation system is determined by the interaction relationship, wherein the test variable is a variable that can be directly measured in the experimental space. In one embodiment, an element can be decomposed into sub-elements that cannot contain each other. The set composed of sub-elements is called a decomposition relationship set, and the relationship between sub-elements is a decomposition relationship; if an element of an entity element represents an element of an attribute element, the two are called a dependency relationship. If the sub-decomposition element of the entity element can also represent the element attribute of the attribute element, the two elements are called a local dependency relationship, otherwise it is a global dependency relationship; if an element of a variable element represents an element of an interaction element or an attribute element, the relationship between the two is called a descriptive relationship.

[0037] In addition, the interaction relationship elements between the entity element and the decomposed sub-elements constitute the epigenetic decomposition relationship; the interaction element and the attribute element and the variable element under the description constitute the epigenetic description relationship; the element relationship of all the description behaviors of the element set is the configuration relationship; if the credibility of element A is lower than the threshold value, element B is considered untrustworthy, otherwise the credibility of B is comprehensively judged, then the relationship between A and B is the threshold relationship, A is the judgment element, B is the target element, then the above relationship set constitutes the inter-element relationship of confidence assessment, and the intersection of these relationships is an empty set. By comprehensively utilizing the relationships of various elements, a confidence assessment model can be constructed, in which the confidence assessment model contains the formal representation of indicators such as attribute information, data information, model characteristics, and predicted data.

[0038] In another embodiment, the test space confidence symbol representation system further classifies key elements in different evaluation processes of the test space confidence evaluation into four types of element sets: element sets: ;Yuanji: ; Confidence set: ;Threshold set: ; Algorithm set: .

[0039] Furthermore, two evaluation algorithms are set in the test space confidence symbol representation system. ,and . It is a confidence assessment algorithm used to implement the confidence assessment results from multi-source information, models or prediction results, which is called the first-level assessment algorithm. It is an evaluation algorithm used to evaluate or calculate multiple confidence levels to a small number of confidence levels, which is called a secondary evaluation algorithm.

[0040] Step S106, using the test space confidence symbol representation system, respectively analyze and process the confidence assessment indicators of each dimension in the confidence assessment indicator set, determine the indicator formal representation model corresponding to each confidence assessment indicator, and integrate the indicator formal representation model to construct a confidence assessment model, wherein the confidence assessment indicator set includes: multi-source information confidence, model confidence and prediction result confidence. In one embodiment, the confidence assessment symbol representation indicator system is defined from three aspects: multi-source information confidence, model confidence, and prediction result confidence, and a total of 6 dimensions of confidence assessment indicators are summarized, and the corresponding confidence assessment model is constructed according to the indicator system.

[0041] Specifically, the confidence of the parameter element is the confidence of the multi-source information; the confidence of the model refers to the confidence of the various characteristics of the simulation system model itself (such as atmospheric density, material parameters, model divergence, transparency, etc.), which mainly evaluates the confidence of the parameters that constitute the model and the interaction within the model that constitutes a single system and the joint action of the model itself under the predicted data output; the confidence of the coupled element refers to the degree of credibility of the simulation system in information interaction, and can also be used as the calculation of the confidence under the interaction between coupled models; the confidence calculation of the physical element (physical system) can also be used as the confidence calculation of the prediction result.

[0042] In summary, in actual academic research and engineering applications, a large number of different types of confidence assessment expressions and definitions will appear, among which frequently appearing similar words include: accuracy, uncertainty, precision, error divergence, etc. The meanings represented by these words are different in different application scenarios. For scholars who do not fully understand the specific calculation method or calculation logic, it is often easy to cause conceptual confusion. The above-mentioned confidence assessment model construction method provided by the embodiment of the present invention combines various confidence assessment indicators commonly used in academia and engineering, divides confidence into 6 dimensions, and defines these dimensions respectively, which can significantly improve the evaluation accuracy.

[0043] Furthermore, based on the symbolic representation system of the test space confidence, the index analysis and processing can be performed for the multi-source information confidence, the model confidence and the prediction result confidence, respectively, to determine the multi-source information confidence formal representation model, the model confidence formal representation model and the prediction result confidence formal representation model, wherein the multi-source information confidence formal representation model is used to evaluate the credibility of the simulation data, physical test data, key factors and prior knowledge input to the model, the model confidence formal representation model is used to evaluate the credibility of the model and the prototype, and the prediction result confidence formal representation model is used to evaluate the credibility of the model output result. In one embodiment, see Figure 2The schematic diagram of the structure of a test space confidence symbol representation system is shown. The embodiment of the present invention also provides an implementation method for constructing a confidence assessment model. For details, see (1) to (3) below:

[0044] (1) Through the experimental space confidence symbolic representation system, the indicator analysis and processing are carried out on the sample point quality and the multi-source information consistency, and the sample point quality formal representation model and the multi-source information consistency formal representation model are determined. The sample point quality formal representation model and the multi-source information consistency formal representation model are combined to determine the multi-source information confidence formal representation model, where the sample point quality is the characteristic measurement of the sample point, and the multi-source information consistency is used to represent the consistency of information from different sources relative to the real information when representing the same indicator information.

[0045] Specifically, when determining the formal representation model of sample point quality, the confidence evaluation objects for multi-source information are various attributes and sample points: an indicator (dimension) of multi-source information confidence can be defined as sample point quality, and the connotation of this quality is the information or feature measurement of the sample point representation model. Assume that the attributes in the meta-set There is a need for sample point quality assessment, and the corresponding assessment method in the algorithm set (first-level assessment algorithm) is used , the sample point quality assessment results (denoted as ) is calculated as:

[0046]

[0047] When there are multiple attributes that need to be evaluated for sample point quality, a secondary evaluation algorithm is used The single-dimensional evaluation results are integrated, and the sample point quality evaluation results (denoted as ) is characterized by:

[0048]

[0049] Furthermore, the connotation of multi-source information consistency is to check the consistency of different information from various sources relative to the real information when the information represents the same indicator information (such as the data of a time series, the input and output information of the same working condition, etc.), such as indicators reflecting data integrity, data loss and data noise; for each multi-source information of the model or system, the first-level algorithm is used to check the consistency of information from different sources relative to the real information. Comprehensive consistency judgment results, combined with threshold conditions Determine whether the consistency judgment is passed. The passing indicator can be a qualitative or quantitative evaluation. Assume that the attributes or subsystem attributes that need to be analyzed for multi-source information consistency are , then the consistency of multi-source information (denoted as ) is characterized by:

[0050]

[0051] (2) Through the test space confidence symbolic representation system, the first model uncertainty and the second model uncertainty are analyzed and processed to determine the formal representation model of the first model uncertainty and the formal representation model of the second model uncertainty. The formal representation model of the first model uncertainty and the formal representation model of the second model uncertainty are combined to determine the model confidence formal representation model, wherein the first model uncertainty is the uncertainty corresponding to the model with changed parameters, and the second model uncertainty is the uncertainty corresponding to the model with unchanged model and changed input.

[0052] Specifically, when evaluating the confidence of a model, the main focus is on the attributes, functions, logic, and mechanisms of the model itself. The prediction results of the model are not within the scope of the model confidence evaluation. The uncertainty of the model is defined to represent the divergence measure of the uncertainty of the model's attributes, mechanisms, and other information on the prediction results.

[0053] For the formulated mechanism model, the uncertain factors that make up the model include its design parameters, measurement parameters, random parameters, black box parameters, etc. The meaning of uncertainty assessment of the mechanism model is to use the first-level assessment algorithm Make these parameters An assessment of the impact of the model's prediction divergence, denoted as model parameter uncertainty (denoted as ), characterized by:

[0054]

[0055] Furthermore, for a model with a constant model and a changing input, when performing simulation calculations, the uncertainty of the input itself will lead to a corresponding uncertainty component in the model output, which is called the input uncertainty of the model. These uncertainties are evaluated by the first-level algorithm. The input parameters of the model are calculated by pulling or experimental design to obtain the uncertainty assessment result, which is recorded as the model input uncertainty (recorded as ) is characterized by:

[0056]

[0057] (3) Through the test space confidence symbolic representation system, the accuracy and 95% confidence indicators are analyzed and processed to determine the accuracy formal representation model and the 95% confidence formal representation model. The accuracy formal representation model and the 95% confidence formal representation model are combined to determine the prediction result confidence formal representation model, where accuracy is used to represent the deviation of the model prediction results and the system prediction results relative to the benchmark data.

[0058] Specifically, the confidence of the prediction results mainly refers to the deviation of the model prediction results and the system prediction results from the benchmark data. The evaluation method of this type of deviation is generally different first-level evaluation algorithms suitable for different data types, such as spectral analysis for periodic time series data and grey correlation method for non-periodic time series data. This type of deviation is recorded as accuracy. , characterized by:

[0059]

[0060] Furthermore, for the classic traditional confidence level in engineering, such as 95% confidence level, it is generally represented by the credibility of the predicted data under the 95% confidence interval. , characterized by:

[0061]

[0062] Based on the above implementation method of constructing a confidence assessment model, the confidence assessment model can cover most of the confidence assessment types and assessment connotations in current academic research and engineering applications. In practical applications, indicators that can be evaluated can be selected according to the characteristics of the actual application object, and it is not necessarily possible to implement the evaluation of 6 indicators for each assessment object.

[0063] See also Figure 3 The flowchart of a confidence evaluation method shown in FIG. 1 mainly includes the following steps S302 to S304:

[0064] Step S302, obtaining the data to be evaluated in the test space to be tested, wherein the data to be evaluated includes: sample point information, model information, and simulation and test data information.

[0065] Step S304, respectively sending the sample point information, model information and simulation and test data information to the pre-built confidence assessment model, determining the multi-source information confidence assessment results, model confidence assessment results and prediction result confidence assessment results, and determining the multi-source information confidence assessment results, model confidence assessment results and prediction result confidence assessment results as the target confidence assessment results in the test space to be tested, wherein the confidence assessment model is constructed based on the above-mentioned confidence assessment model construction method. In one embodiment, when the uncertainty of the parameters of the solid-liquid rocket engine is evaluated by the confidence assessment model and the model confidence assessment results are determined, taking the uncertainty assessment based on orthogonal test design and variance analysis as an example, it is necessary to set different levels and modify the model parameters, perform variance analysis on the obtained test results, and finally obtain the uncertainty of the model. After conducting the test according to the orthogonal test table, the sum of the squares of the total deviation can be obtained, and then the F ratio of each parameter is calculated, and finally the uncertainty of each parameter and the total uncertainty evaluation result are obtained, wherein the calculation formula of the F ratio is:

[0066]

[0067] in is the sum of squares of the test factor deviations, is the effect of the test factor or prediction result when the level changes under the model working conditions, is the corresponding degree of freedom.

[0068] Furthermore, in order to meet the design and use requirements of solid-liquid rocket engines and verify the uncertainty quantification evaluation method based on orthogonal experimental design and variance analysis, by analyzing the impact of the change of experimental factor levels and their interactions on thrust, the steps for quantitatively evaluating the simulation uncertainty of the power subsystem include the following (a) to (c):

[0069] (a) Selection of experimental factors, variation levels and table design: First, it is necessary to select appropriate experimental factors. The analysis of the power subsystem simulation program is mainly focused on the main component model, namely the combustion chamber model. Model parameters such as combustion efficiency, overall efficiency, ablation rate, start ablation time, combustion efficiency, and two-stage combustion rate formula parameters can only be set based on relevant literature and the experience of professionals, thus introducing uncertainty. These are selected as experimental factors to obtain the simulation test factor and level table, and then proceed to the next step of header design. If there is data with a coupling relationship, additional experiments are required.

[0070] (b) Conducting experiments and analyzing results: After obtaining the uncertainty analysis test table, simulation tests are performed based on the orthogonal table to obtain the results of variance analysis, thereby obtaining the square of the deviation of each factor and the degrees of freedom of the response error.

[0071] (c) Model uncertainty assessment: The uncertainty is calculated based on the results of variance analysis. After obtaining the thrust uncertainty and uncertainty band of each component, the quantitative assessment results of the uncertainty of the time series are integrated into the overall uncertainty. Finally, the average value of the steady-state thrust of the first stage of the physical experiment results is taken to calculate the relative uncertainty. The quantitative assessment results of the simulation uncertainty are shown in Table 1.

[0072] Table 1 Uncertainty evaluation results

[0073]

[0074] In the table, the uncertainties corresponding to the uncertainty assessment order from A to H are: combustion efficiency A, overall efficiency B, ablation rate C, start ablation time D, first-stage burning rate formula parameter E, first-stage burning rate formula parameter F, second-stage burning rate formula parameter G, and second-stage burning rate formula parameter H.

[0075] It can be observed that the combined uncertainty of the thrust response is about 3.9%.

[0076] The confidence evaluation method provided by the embodiment of the present invention can evaluate the information of the experimental space using the confidence evaluation model, thereby significantly improving the evaluation accuracy.

[0077] In summary, starting from the formalized evaluation model, the present invention constructs a confidence evaluation model with three dimensions and six indicators and its symbolic representation system, and uses two different algorithm sets to evaluate the confidence of the test space generated by the prototype system engineering. It can cover the existing confidence evaluation needs and evaluation indicators in engineering and academia, has clear engineering significance, and the calculation results have reliable reference value for system design. At the same time, since the present invention only relies on multi-source information, including physical test data, simulation experiment data, proxy model information, finite element model information, etc., and does not involve specific physical operations and production processing, there is no need for a model entity. Furthermore, the present invention can use the confidence evaluation model and the confidence symbolic representation indicator system to perform likelihood estimation of the test space in any complex system involving test space design to support system design.

[0078] Among the three dimensions and six indicators constructed, multi-source information confidence evaluates the credibility of various types of information in the system. This type of confidence evaluation results provide technical support and guidance for the construction of the test space and the direction of test development; model confidence evaluates the influence of various model parameters and input variables in the test space model library on the model divergence. This type of confidence evaluation results provides technical support for the deployment, integration and use of the test space model; prediction result confidence evaluates the credibility of the system functional effect. This type of confidence evaluation results provides technical support for the use of the test space.

[0079] For the confidence assessment model construction method provided in the above embodiment, the embodiment of the present invention provides a confidence assessment model construction device, see Figure 4 A schematic diagram of the structure of a confidence assessment model building device is shown, and the device includes the following parts:

[0080] The data acquisition module 402 acquires a preset element set, element relationship and confidence evaluation index set in the test space, wherein the element set includes: a meta set, a confidence set, an algorithm set and a threshold set, and the element relationship includes: equivalence, inclusion and true inclusion;

[0081] The interactive relationship analysis module 404 performs interactive relationship analysis on the element set by using the element relationship through a preset formal evaluation model, determines the interactive relationship between each element and the test variable, and determines the test space confidence symbol representation system by using the interactive relationship, wherein the test variable is a variable that can be directly measured in the experimental space;

[0082] The model training module 406 uses the test space confidence symbol representation system to perform indicator analysis and processing on the confidence assessment indicators of each dimension in the confidence assessment indicator set, determines the indicator formal representation model corresponding to each confidence assessment indicator, and integrates the indicator formal representation model to construct a confidence assessment model.

[0083] The above-mentioned confidence assessment model construction device provided in the embodiment of the present application can significantly improve the assessment accuracy.

[0084] In one embodiment, the confidence assessment indicator set includes: multi-source information confidence, model confidence and prediction result confidence. When using the test space confidence symbol representation system to perform indicator analysis and processing on the confidence assessment indicators of each dimension in the confidence assessment indicator set, and determining the indicator formalization representation model corresponding to each confidence assessment indicator, the above-mentioned model training module 406 is also used to: based on the test space confidence symbol representation system, perform indicator analysis and processing on the multi-source information confidence, model confidence and prediction result confidence, and determine the multi-source information confidence formalization representation model, model confidence formalization representation model and prediction result confidence formalization representation model, wherein the multi-source information confidence formalization representation model is used to evaluate the credibility of the simulation data, physical test data, key factors and prior knowledge of the model input, the model confidence formalization representation model is used to evaluate the credibility of the model and prototype, and the prediction result confidence formalization representation model is used to evaluate the credibility of the model output result.

[0085] In one embodiment, when performing an indicator analysis and processing based on the experimental space confidence symbolic representation system to determine the step of formalizing the multi-source information confidence representation model, the above-mentioned model training module 406 is also used to: perform indicator analysis and processing on sample point quality and multi-source information consistency through the experimental space confidence symbolic representation system to determine the sample point quality formalization representation model and the multi-source information consistency formalization representation model, wherein the sample point quality is a characteristic measure of the sample point, and the multi-source information consistency is used to represent the consistency of information from different sources relative to the real information when representing the same indicator information; combine the sample point quality formalization representation model and the multi-source information consistency formalization representation model to determine the multi-source information confidence formalization representation model.

[0086] In one embodiment, when performing the step of performing index analysis and processing on model confidence based on the experimental space confidence symbolic representation system to determine the model confidence formal representation model, the above-mentioned model training module 406 is also used to: perform index analysis and processing on the first model uncertainty and the second model uncertainty through the experimental space confidence symbolic representation system to determine the first model uncertainty formal representation model and the second model uncertainty formal representation model, wherein the first model uncertainty is the uncertainty corresponding to the model with changed parameters, and the second model uncertainty is the uncertainty corresponding to the model with unchanged model and changed input; combine the first model uncertainty formal representation model and the second model uncertainty formal representation model to determine the model confidence formal representation model.

[0087] In one embodiment, when performing the step of performing indicator analysis and processing on the prediction result confidence based on the experimental space confidence symbolic representation system, and determining the prediction result confidence formal representation model, the above-mentioned model training module 406 is also used to: perform indicator analysis and processing on accuracy and 95% confidence through the experimental space confidence symbolic representation system, and determine the accuracy formal representation model and the 95% confidence formal representation model, wherein the accuracy is used to represent the deviation of the model prediction result and the system prediction result relative to the benchmark data; combine the accuracy formal representation model and the 95% confidence formal representation model to determine the prediction result confidence formal representation model.

[0088] For the confidence evaluation method provided in the above embodiment, the embodiment of the present invention provides a confidence evaluation device, see Figure 5 A schematic diagram of the structure of a confidence assessment device is shown, the device comprises the following parts:

[0089] The data acquisition module 502 acquires the data to be evaluated in the test space to be tested, wherein the data to be evaluated includes: sample point information, model information, and simulation and test data information;

[0090] The confidence assessment module 504 sends the sample point information, model information, and simulation and test data information to a pre-built confidence assessment model, respectively, determines the multi-source information confidence assessment results, the model confidence assessment results, and the prediction result confidence assessment results, and determines the multi-source information confidence assessment results, the model confidence assessment results, and the prediction result confidence assessment results as the target confidence assessment results in the test space to be tested; wherein the confidence assessment model is constructed based on the above-mentioned confidence assessment model construction method.

[0091] The above-mentioned confidence assessment device provided in the embodiment of the present application can use the confidence assessment model to evaluate the information of the experimental space, thereby significantly improving the assessment accuracy.

[0092] The device provided in the embodiment of the present invention has the same implementation principle and technical effects as those of the aforementioned method embodiment. For the sake of brief description, for matters not mentioned in the device embodiment, reference may be made to the corresponding contents in the aforementioned method embodiment.

[0093] An embodiment of the present invention provides a server. Specifically, the server includes a processor and a storage device. The storage device stores a computer program, and when the computer program is executed by the processor, it executes the method described in any one of the above-mentioned embodiments.

[0094] Figure 6 A structural diagram of a server provided in an embodiment of the present invention, the server 100 includes: a processor 60, a memory 61, a bus 62 and a communication interface 63, wherein the processor 60, the communication interface 63 and the memory 61 are connected via the bus 62; the processor 60 is used to execute an executable module stored in the memory 61, such as a computer program.

[0095] The memory 61 may include a high-speed random access memory (RAM), and may also include a non-volatile memory, such as at least one disk memory. The communication connection between the system network element and at least one other network element is realized through at least one communication interface 63 (which may be wired or wireless), and the Internet, wide area network, local area network, metropolitan area network, etc. may be used.

[0096] The bus 62 may be an ISA bus, a PCI bus, or an EISA bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 6 Only one bidirectional arrow is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0097] Among them, the memory 61 is used to store programs, and the processor 60 executes the program after receiving the execution instruction. The method executed by the device for flow process definition disclosed in any embodiment of the above-mentioned embodiment of the present invention can be applied to the processor 60 or implemented by the processor 60.

[0098] The processor 60 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit or software instructions in the processor 60. The above processor 60 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present invention can be implemented or executed. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor. The steps of the method disclosed in the embodiment of the present invention can be directly embodied as a hardware decoding processor to execute, or can be executed by a combination of hardware and software modules in the decoding processor. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 61, and the processor 60 reads the information in the memory 61 and completes the steps of the above method in combination with its hardware.

[0099] The computer program product of the readable storage medium provided in the embodiment of the present invention includes a computer-readable storage medium storing program code, and the instructions included in the program code can be used to execute the methods described in the previous method embodiments. The specific implementation can be referred to the previous method embodiments, which will not be repeated here.

[0100] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0101] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A confidence assessment model construction method, characterized in that: The method is applied to the design field of complex system test space in prototype system engineering, and the method comprises: Obtain preset element sets, element relationships, and confidence evaluation index sets in the test space, wherein the element sets include: element sets, confidence sets, algorithm sets, and threshold sets, and the element relationships include: equivalence, inclusion, and true inclusion. The speed and thrust of the prototype system during operation are encapsulated as states, and the states between different systems or the sets of different states of the same system are taken as element sets. In the element sets, the set of simulation modules is taken as entity elements, and the set of action relationships between simulation modules is taken as interaction elements. Inertia, conductivity, shape, color, and ablation rate in a solid-liquid rocket engine thrust simulation program are taken as attributes of the simulation object, and the set of attributes of the simulation object is taken as attribute elements. By using the preset formal evaluation model and the element relationship, the element set is subjected to interactive relationship analysis and processing, the interactive relationship between each element and the test variable is determined, and the test space confidence symbol representation system is determined using the interactive relationship, wherein the test variable is a variable that can be directly measured in the experimental space; Using the test space confidence symbol representation system, respectively perform indicator analysis and processing on the confidence assessment indicators of each dimension in the confidence assessment indicator set, determine the indicator formalization representation model corresponding to each of the confidence assessment indicators, and integrate the indicator formalization representation models to construct a confidence assessment model; Wherein, the confidence assessment index set includes: multi-source information confidence, model confidence and prediction result confidence, and the step of using the test space confidence symbol representation system to perform indicator analysis and processing on the confidence assessment indicators of each dimension in the confidence assessment index set respectively, and determining the indicator formalization representation model corresponding to each of the confidence assessment indicators includes: based on the test space confidence symbol representation system, respectively performing indicator analysis and processing on the multi-source information confidence, the model confidence and the prediction result confidence, and determining the multi-source information confidence formalization representation model, the model confidence formalization representation model and the prediction result confidence formalization representation model, wherein the multi-source information confidence formalization representation model is used to evaluate the credibility of the simulation data, physical test data, key factors and prior knowledge of the model input, the model confidence formalization representation model is used to evaluate the credibility of the model and prototype, and the prediction result confidence formalization representation model is used to evaluate the credibility of the model output result; Among them, based on the test space confidence symbolic representation system, the index analysis and processing is performed on the multi-source information confidence, and the step of determining the multi-source information confidence formal representation model includes: through the test space confidence symbolic representation system, the index analysis and processing is performed on the sample point quality and the multi-source information consistency, and the sample point quality formal representation model and the multi-source information consistency formal representation model are determined, wherein the sample point quality is a characteristic measurement of the sample point, and the multi-source information consistency is used to represent the consistency of information from different sources relative to the real information when representing the same indicator information; the sample point quality formal representation model and the multi-source information consistency formal representation model are combined to determine the multi-source information confidence formal representation model; Among them, based on the test space confidence symbolic representation system, the model confidence is subjected to index analysis and processing, and the step of determining the model confidence formal representation model includes: through the test space confidence symbolic representation system, the first model uncertainty and the second model uncertainty are subjected to index analysis and processing, and the first model uncertainty formal representation model and the second model uncertainty formal representation model are determined, wherein the first model uncertainty is the uncertainty corresponding to the model with parameter changes, and the second model uncertainty is the uncertainty corresponding to the model with unchanged model and input changes; the first model uncertainty formal representation model and the second model uncertainty formal representation model are combined to determine the model confidence formal representation model.

2. The confidence assessment model construction method according to claim 1, characterized in that: Based on the test space confidence symbol representation system, the steps of performing index analysis and processing on the prediction result confidence and determining the prediction result confidence formal representation model include: Through the test space confidence symbol representation system, the accuracy and 95% confidence are analyzed and processed to determine the accuracy formalization representation model and the 95% confidence formalization representation model, wherein the accuracy is used to represent the deviation of the model prediction results and the system prediction results relative to the benchmark data; The accuracy formalization characterization model and the 95% confidence formalization characterization model are combined to determine the prediction result confidence formalization characterization model.

3. A confidence assessment method, characterized in that: The method comprises: Acquire the data to be evaluated in the test space to be tested, wherein the data to be evaluated includes: sample point information, model information, and simulation and test data information; The sample point information, the model information and the simulation and test data information are respectively sent to a pre-built confidence assessment model to determine a multi-source information confidence assessment result, a model confidence assessment result and a prediction result confidence assessment result, and the multi-source information confidence assessment result, the model confidence assessment result and the prediction result confidence assessment result are determined as target confidence assessment results in the test space to be tested; Wherein, the confidence assessment model is constructed based on the confidence assessment model construction method described in any one of claims 1-2.

4. A confidence assessment model construction device, characterized in that: The device is applied to the design field of complex system test space in prototype system engineering, and the device comprises: A data acquisition module is used to acquire a preset element set, element relationship and confidence evaluation index set in the test space, wherein the element set includes: a meta set, a confidence set, an algorithm set and a threshold set, and the element relationship includes: equivalence, inclusion and true inclusion. The speed and thrust of the prototype system during operation are encapsulated as states, and the states between different systems or the set consisting of different states of the same system is taken as a meta set. In the meta set, the set consisting of simulation modules is taken as an entity element, and the set consisting of the action relationship between simulation modules is taken as an interaction element. Inertia, conductivity, shape, color and ablation rate in a solid-liquid rocket engine thrust simulation program are taken as attributes of the simulation object, and the set consisting of the attributes of the simulation object is taken as an attribute element. The interactive relationship analysis module uses the element relationship to perform interactive relationship analysis on the element set by presetting a formalized evaluation model, determines the interactive relationship between each element and the test variable, and determines the test space confidence symbol representation system using the interactive relationship, wherein the test variable is a variable that can be directly measured in the experimental space; The model training module uses the test space confidence symbol representation system to perform indicator analysis and processing on the confidence assessment indicators of each dimension in the confidence assessment indicator set, determines the indicator formalization representation model corresponding to each confidence assessment indicator, and integrates the indicator formalization representation model to construct a confidence assessment model; Wherein, the confidence assessment index set includes: multi-source information confidence, model confidence and prediction result confidence, and the step of using the test space confidence symbol representation system to perform indicator analysis and processing on the confidence assessment indicators of each dimension in the confidence assessment index set respectively, and determining the indicator formalization representation model corresponding to each of the confidence assessment indicators includes: based on the test space confidence symbol representation system, respectively performing indicator analysis and processing on the multi-source information confidence, the model confidence and the prediction result confidence, and determining the multi-source information confidence formalization representation model, the model confidence formalization representation model and the prediction result confidence formalization representation model, wherein the multi-source information confidence formalization representation model is used to evaluate the credibility of the simulation data, physical test data, key factors and prior knowledge of the model input, the model confidence formalization representation model is used to evaluate the credibility of the model and prototype, and the prediction result confidence formalization representation model is used to evaluate the credibility of the model output result; Among them, based on the test space confidence symbolic representation system, the index analysis and processing is performed on the multi-source information confidence, and the step of determining the multi-source information confidence formal representation model includes: through the test space confidence symbolic representation system, the index analysis and processing is performed on the sample point quality and the multi-source information consistency, and the sample point quality formal representation model and the multi-source information consistency formal representation model are determined, wherein the sample point quality is a characteristic measurement of the sample point, and the multi-source information consistency is used to represent the consistency of information from different sources relative to the real information when representing the same indicator information; the sample point quality formal representation model and the multi-source information consistency formal representation model are combined to determine the multi-source information confidence formal representation model; Among them, based on the test space confidence symbolic representation system, the model confidence is subjected to index analysis and processing, and the step of determining the model confidence formal representation model includes: through the test space confidence symbolic representation system, the first model uncertainty and the second model uncertainty are subjected to index analysis and processing, and the first model uncertainty formal representation model and the second model uncertainty formal representation model are determined, wherein the first model uncertainty is the uncertainty corresponding to the model with parameter changes, and the second model uncertainty is the uncertainty corresponding to the model with unchanged model and input changes; the first model uncertainty formal representation model and the second model uncertainty formal representation model are combined to determine the model confidence formal representation model.

5. A confidence assessment device, characterized in that: The device comprises: A data acquisition module is used to obtain the data to be evaluated in the test space to be tested, wherein the data to be evaluated includes: sample point information, model information, and simulation and test data information; A confidence assessment module sends the sample point information, the model information, and the simulation and test data information to a pre-built confidence assessment model, respectively, determines a multi-source information confidence assessment result, a model confidence assessment result, and a prediction result confidence assessment result, and determines the multi-source information confidence assessment result, the model confidence assessment result, and the prediction result confidence assessment result as a target confidence assessment result in a test space to be tested; Wherein, the confidence assessment model is constructed based on the confidence assessment model construction method described in any one of claims 1-2.

6. A server, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer executable instructions that can be executed by the processor, and the processor executes the computer executable instructions to implement the method according to any one of claims 1 to 2.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, and when the computer-executable instructions are called and executed by a processor, the computer-executable instructions prompt the processor to implement the method according to any one of claims 1 to 2.

Images