A model verification method and apparatus based on measured data

Abstract

This invention discloses a model verification method and apparatus based on measured data. The method includes: acquiring a measured dataset; processing the measured dataset to obtain a first dataset; processing the first dataset to obtain a second dataset and a third dataset; processing the second dataset to obtain a fourth dataset; and processing the third and fourth datasets to obtain model evaluation result information. This invention performs singular value removal, time alignment, spatial alignment, and parameter sensitivity analysis on the acquired measured data from different platforms to obtain platform performance parameter data and test data, which are used for verifying and evaluating platform simulation models, thus improving the realism of the simulation models.

Description

Technical Field

[0001] This invention relates to the field of model simulation evaluation technology, and in particular to a model verification method and apparatus based on measured data. Background Technology

[0002] The realism of equipment simulation models directly affects the credibility of equipment combat simulation results. Therefore, how to verify simulation models has become a key research focus. Verification of simulation models is achieved by converting field measurement data into data that can be processed by the simulation model and injecting it into the simulation model's output. The results are then compared with the field measurement data.

[0003] Verification, validation, and acceptance of simulation systems are fundamental to their confidence level and are essential processes in the development of any simulation system. Currently, research on the reliability of simulation systems in my country is still in its early stages and lags behind international advanced levels. Effectively utilizing measured data to verify simulation models, achieving consistency checks between internal and external fields, and improving the realism of simulation models are pressing technical problems that need to be addressed. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a model verification method and device based on measured data, which effectively utilizes measured data to verify the simulation model, realizes the consistency verification of internal and external fields, improves the realism of the simulation model, and realizes the potential value of equipment test data and deepens the connotation of equipment test and evaluation.

[0005] To address the aforementioned technical problems, a first aspect of the present invention discloses a model verification method based on measured data, the method comprising:

[0006] S1, Obtain the actual test dataset;

[0007] S2, process the measured dataset to obtain the first dataset;

[0008] S3, process the first dataset to obtain the second dataset and the third dataset;

[0009] S4, process the second dataset to obtain the fourth dataset;

[0010] S5, process the third and fourth datasets to obtain model evaluation results.

[0011] As an optional implementation, in the first aspect of the present invention, processing the measured dataset to obtain a first dataset includes:

[0012] S21, The measured dataset is processed using a data elimination model to obtain a first preprocessed dataset;

[0013] The data removal model includes a first data removal model and a second data removal model;

[0014] S22, using a data validation model, the first preprocessed dataset is processed to obtain a second preprocessed dataset;

[0015] The data verification model includes a first data verification model and a second data verification model;

[0016] S23, Perform validation processing on the second preprocessed dataset to obtain the third preprocessed dataset;

[0017] S24, perform error processing on the third preprocessed dataset to obtain the fourth preprocessed dataset;

[0018] S25, perform time alignment processing on the fourth preprocessed dataset to obtain the fifth preprocessed dataset;

[0019] S26, Spatial alignment processing is performed on the fifth preprocessed dataset to obtain the first dataset.

[0020] As an optional implementation, in the first aspect of the present invention, the step of processing the measured dataset using a data elimination model to obtain a first preprocessed dataset includes:

[0021] S211, Obtain the data sample size value of the measured dataset;

[0022] S212, determine whether the data sample size is greater than a first threshold, and obtain a first determination result;

[0023] S213, when the first judgment result is yes, execute S214;

[0024] If the first judgment result is negative, execute S1;

[0025] S214, determine whether the data sample size is less than the second threshold, and obtain the second determination result;

[0026] When the second judgment result is yes, the measured dataset is processed using the first data elimination model to obtain the first preprocessed dataset;

[0027] The expression for the first data removal model is:

[0028] ;

[0029] in, This represents the data sample size of the measured dataset; Represents the first in the measured dataset One measured data point; Indicates the index of the measured data; S represents the mean of the measured dataset; S represents the standard deviation of the measured data.

[0030] When the second judgment result is negative, the measured dataset is processed using the second data elimination model to obtain the first preprocessed dataset;

[0031] The expression for the second data removal model is:

[0032] ;

[0033] in, This represents the data sample size of the measured dataset; Represents the first in the measured dataset One measured data point; Indicates the index of the measured data; This represents the average value of the measured dataset; Indicates the significance level; Indicates the first relation coefficient; The standard deviation of the measured dataset is represented by .

[0034] As an optional implementation, in the first aspect of the present invention, the step of processing the first preprocessed dataset using a data validation model to obtain a second preprocessed dataset includes:

[0035] S221, Obtain the data sample size value of the first preprocessed dataset;

[0036] S222, determine whether the data sample size is greater than the third threshold, and obtain the third judgment result;

[0037] S223, when the third judgment result is yes, execute S224;

[0038] If the result of the third judgment is negative, execute S1;

[0039] S224, determine whether the data sample size is less than the fourth threshold, and obtain the fourth judgment result;

[0040] When the fourth judgment result is yes, the first preprocessed dataset is processed using the first data verification model to obtain the second preprocessed dataset;

[0041] The expression for the first data verification model is:

[0042] ;

[0043] in, This represents the sample order statistics for the first preprocessed dataset; This represents the number of data samples in the first preprocessed dataset; Indicates the first The significance level of the first preprocessed data; Indicates the significance level quantile; Indicates the first relation coefficient;

[0044] When the fourth judgment result is negative, the first preprocessed dataset is processed using the second data verification model to obtain the second preprocessed dataset;

[0045] The expression for the second data verification model is:

[0046]

[0047] in, This represents the sample order statistics for the first preprocessed dataset; This represents the number of data samples in the first preprocessed dataset; Indicates the first The significance level of the first preprocessed data; This indicates the rejection field for the second data verification.

[0048] As an optional implementation, in a first aspect of the present invention, the step of performing verification processing on the second preprocessed dataset to obtain a third preprocessed dataset includes:

[0049] S231, perform autocorrelation processing on each element of the second preprocessed dataset to obtain the correlation dataset;

[0050] The autocorrelation processing expression is:

[0051] ;

[0052] in, The second preprocessed dataset represents the first... The second preprocessed data;

[0053] S232, perform estimation processing on each element of the relevant processing dataset to obtain a set of relevant data estimates;

[0054] The estimation processing expression is:

[0055] ;

[0056] in, It is expressed as the autocorrelation coefficient of white noise; Represents the first in the relevant dataset One relevant data point; This indicates the number of elements in the relevant dataset;

[0057] S233, the estimated value set of the relevant data is judged and processed to obtain the third preprocessed dataset.

[0058] As an optional implementation, in a first aspect of the present invention, processing the first dataset to obtain a second dataset and a third dataset includes:

[0059] S31, parse the first dataset to obtain the test parameter dataset and the test result dataset;

[0060] S32, using a sensitivity analysis model, the test parameter dataset and the test result dataset are processed to obtain a second dataset and a third dataset;

[0061] The sensitivity analysis model expression is:

[0062] ;

[0063] ;

[0064] in, Indicates the first mutual information value; Indicates the second mutual information value; This represents the second dataset; This refers to the third dataset; Represents random events for The probability of; Represents random events for The probability of; Represents random events and random events The joint probability.

[0065] As an optional implementation, in the first aspect of the present invention, processing the second dataset to obtain a fourth dataset includes:

[0066] S41, process the second dataset to obtain the second optimized dataset;

[0067] S42, using the parameter control model, iteratively process the second optimized dataset to obtain the fourth dataset;

[0068] The expression for the parameter control model is:

[0069] ;

[0070] ;

[0071] in, Indicates the first Parameters at the next iteration The value; Indicates the first Parameters at the next iteration The value; Indicates the learning rate; Represents the prediction function; This indicates that the prediction function has respect to the parameters. In the Partial derivative at the next iteration; Indicates the first Secondary parameter update step size; Indicates the learning rate; Represents the loss function; Represents the loss function Regarding parameters In the The gradient at the next iteration.

[0072] A second aspect of this invention discloses a model verification device based on measured data, the device comprising:

[0073] The data acquisition module is used to acquire the actual test dataset;

[0074] The first processing module is used to process the measured dataset to obtain the first dataset;

[0075] The second processing module is used to process the first dataset to obtain the second dataset and the third dataset;

[0076] The third processing module is used to process the second dataset to obtain the fourth dataset;

[0077] The fourth processing module is used to process the third and fourth datasets to obtain model evaluation results.

[0078] A third aspect of the present invention discloses a model verification device based on measured data, the device comprising:

[0079] Memory containing executable program code;

[0080] A processor coupled to the memory;

[0081] The processor calls the executable program code stored in the memory to execute some or all of the steps in the model verification method based on measured data disclosed in the first aspect of the present invention.

[0082] The fourth aspect of the present invention discloses a computer-readable storage medium storing computer instructions, which, when invoked, execute some or all of the steps in the model verification method based on measured data disclosed in the first aspect of the present invention.

[0083] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0084] In this embodiment of the invention, the platform performance parameter data and measured data are obtained by performing singular value removal, time alignment, spatial alignment and parameter sensitivity analysis on the acquired measured data from different platforms. These are used for the verification and evaluation of the platform simulation model, thereby improving the realism of the simulation model. Attached Figure Description

[0085] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0086] Figure 1 This is a schematic diagram of a simulation system provided in an embodiment of the present invention;

[0087] Figure 2 This is a flowchart illustrating a model verification method based on measured data disclosed in an embodiment of the present invention;

[0088] Figure 3 This is a schematic diagram of a model verification device based on measured data disclosed in an embodiment of the present invention;

[0089] Figure 4 This is a schematic diagram of another model verification device based on measured data disclosed in an embodiment of the present invention. Detailed Implementation

[0090] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0091] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this invention are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, apparatus, product, or device that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or devices.

[0092] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0093] In this application, the term "exemplary" is used to mean "used as an example, illustration, or description." Any embodiment described as "exemplary" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use this application. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that this application can be made without using these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of this application with unnecessary detail. Therefore, this application is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0094] It should be noted that since the method in this application embodiment is executed in a computer device, the processing objects of each computer device exist in the form of data or information, such as time, which is essentially time information. It is understood that if size, quantity, position, etc. are mentioned in subsequent embodiments, they are all corresponding data that exist so that the computer device can process them. Specific details will not be elaborated here.

[0095] It should be noted that the artificial intelligence-related technologies that may be involved in this application will be briefly described. Artificial intelligence (AI) is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results. In other words, artificial intelligence is a comprehensive technology in computer science that attempts to understand the essence of intelligence and produce a new kind of intelligent machine that can react in a way similar to human intelligence. Artificial intelligence is the study of the design principles and implementation methods of various intelligent machines, enabling machines to have the functions of perception, reasoning, and decision-making.

[0096] Artificial intelligence (AI) is a comprehensive discipline encompassing a wide range of fields, including both hardware and software technologies. Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies primarily include computer vision, speech processing, natural language processing, and machine learning / deep learning.

[0097] Computer vision (CV) is a science that studies how to enable machines to "see." More specifically, it refers to machine vision, which uses cameras and computers to replace human eyes in recognizing and measuring targets, and then performs image processing to create images more suitable for human observation or transmission to instruments. As a scientific discipline, computer vision studies related theories and technologies, attempting to build artificial intelligence systems capable of extracting information from images or multidimensional data. Computer vision technologies typically include image processing, image recognition, image semantic understanding, image retrieval, OCR, video processing, video semantic understanding, video content / behavior recognition, 3D object reconstruction, 3D technology, virtual reality, augmented reality, simultaneous localization and mapping (SLAM), and common biometric recognition technologies such as facial recognition and fingerprint recognition.

[0098] Monomodal information refers to data of only one type, such as text, images, audio, video, or electromagnetic signals. Multimodal information refers to data that includes at least two types of monomodal information. Furthermore, multimodal information is suitable for complex tasks that require the integration of multiple information sources, such as sentiment analysis, robot interaction, and autonomous driving. By integrating information from multiple modalities, higher performance and accuracy can usually be achieved in these tasks.

[0099] Large models refer to artificial neural network models with a very large number of parameters. In the field of artificial intelligence, large models typically refer to models with hundreds of millions to trillions of parameters. These models usually need to be trained on large-scale datasets and require a significant amount of computing resources for optimization and tuning. Large models are commonly used to solve complex tasks such as natural language processing, computer vision, and speech recognition. Generative AI is a type of AI that can create new content and ideas, including dialogues, stories, images, videos, and music. In this embodiment, the large model can be a language model of the scale of ChatGPT, BERT, XLNet, Zhipu model, Claude, Moonshot-AI model, ChatGLM model, Qianyitongwen model, MiniMax model, Xinghuo model, Llama model, 360GPT model, Qwen model, Baichuan model, Yunque model, vivoLM model, and Wenxin Yiyan, etc., and this embodiment does not limit the scope of the large model.

[0100] This application provides a model verification method, system, device, computer equipment, and computer-readable storage medium based on measured data, which will be described in detail below.

[0101] Please see Figure 1 , Figure 1 This is a schematic diagram of a simulation system provided in an embodiment of this application. The system may include a computer device 100, which integrates a model verification device based on measured data, such as... Figure 1 Computer equipment in the country.

[0102] In this embodiment, the computer device 100 can be a standalone server, a server network, or a server cluster. For example, the computer device 100 described in this embodiment includes, but is not limited to, a computer, a network host, a single network server, a set of multiple network servers, or a cloud server composed of multiple servers. The cloud server is composed of a large number of computers or network servers based on cloud computing.

[0103] It is understood that the computer device 100 used in the embodiments of this application can be a device that includes both receiving and transmitting hardware, that is, a device having receiving and transmitting hardware capable of performing bidirectional communication on a bidirectional communication link. Such a device may include: cellular or other communication devices having a single-line display, a multi-line display, or a cellular or other communication device without a multi-line display. Specifically, the computer device 100 may be a desktop terminal or a mobile terminal, and may also be one of a mobile phone, tablet computer, laptop computer, etc.

[0104] Those skilled in the art will understand that Figure 1 The application environment shown is merely one application scenario of the solution in this application and does not constitute a limitation on the application scenario of the solution in this application. Other application environments may include those that are more specific to this application. Figure 1 The number of computer devices shown is more or less, for example Figure 1 Only one computer device is shown in the diagram. It is understood that the system may also include one or more other services, which are not limited here.

[0105] In addition, such as Figure 1 As shown, the simulation system may also include a memory 200 for storing processing result data and remote sensing image sample data, such as evaluation result data.

[0106] It should be noted that, Figure 1 The schematic diagram of the simulation system shown is merely an example. The simulation system and scenario described in this application are for the purpose of more clearly illustrating the technical solutions of this application and do not constitute a limitation on the technical solutions provided in this application. As those skilled in the art will know, with the evolution of simulation systems and the emergence of new business scenarios, the technical solutions provided in this application are also applicable to similar technical problems.

[0107] This invention discloses a model verification method and apparatus based on measured data. The method involves processing measured data from different platforms by performing singular value removal, time alignment, spatial alignment, and parameter sensitivity analysis to obtain platform performance parameter data. This data is then used for verifying and evaluating the platform simulation model, thereby improving the realism of the simulation model. Detailed explanations follow.

[0108] Example 1

[0109] Please see Figure 2 , Figure 2 This is a flowchart illustrating a model verification method based on measured data disclosed in an embodiment of the present invention. Figure 2 The model verification method based on measured data described herein is applicable to simulation systems, such as local servers or cloud servers used in simulation systems; however, this embodiment of the invention is not limited to such applications. Figure 1 As shown, this model validation method based on measured data may include the following operations:

[0110] S1, Obtain the actual test dataset;

[0111] It should be noted that the measured dataset refers to measured data from different platforms, which is used to verify the simulation model;

[0112] S2, process the measured dataset to obtain the first dataset;

[0113] It should be noted that the measured dataset needs to undergo singular value removal, normality test, independence test, error processing, time alignment, and spatial transformation data preprocessing to obtain a first dataset that is consistent with the simulation data format, units, and coordinate system.

[0114] S3, process the first dataset to obtain the second dataset and the third dataset;

[0115] S4, process the second dataset to obtain the fourth dataset;

[0116] S5, process the third and fourth datasets to obtain model evaluation results.

[0117] As can be seen, by implementing the model verification method based on measured data described in the embodiments of the present invention, the obtained measured data from different platforms are processed by singular value removal, time alignment, spatial alignment, and parameter sensitivity analysis to obtain the third dataset including platform performance parameter data and the fourth dataset of measured results. The platform simulation model is verified and evaluated using the third dataset and the fourth dataset, thereby improving the realism of the simulation model.

[0118] Optionally, in step S2 above, processing the measured dataset to obtain the first dataset includes:

[0119] S21, The measured dataset is processed using a data elimination model to obtain a first preprocessed dataset;

[0120] The data removal model includes a first data removal model and a second data removal model;

[0121] S22, using a data validation model, the first preprocessed dataset is processed to obtain a second preprocessed dataset;

[0122] The data verification model includes a first data verification model and a second data verification model;

[0123] S23, Perform validation processing on the second preprocessed dataset to obtain the third preprocessed dataset;

[0124] S24, perform error processing on the third preprocessed dataset to obtain the fourth preprocessed dataset;

[0125] It should be noted that the error processing mentioned refers to the use of outlier processing, missing value processing, standardization and normalization processing, data cleaning processing, and data smoothing processing to reduce random errors or eliminate systematic errors and improve data quality and reliability. This embodiment does not limit this.

[0126] S25, perform time alignment processing on the fourth preprocessed dataset to obtain the fifth preprocessed dataset;

[0127] S26, Spatial alignment processing is performed on the fifth preprocessed dataset to obtain the first dataset;

[0128] It should be noted that the spatial alignment process refers to unifying the fifth preprocessed dataset of multi-frequency device data into a coordinate system with the fusion center platform as the origin through coordinate transformation. The coordinate systems involved include polar coordinates, rectangular coordinates, geodetic coordinates, and grid coordinates. This embodiment does not limit the scope of the coordinate system.

[0129] It should be noted that the measured datasets are from different platforms and differ from the simulation data format, units, coordinate system, and other aspects required by the target model. Therefore, data preprocessing such as format conversion, coordinate transformation, interpolation extrapolation, and spatiotemporal consistency calculation must be performed.

[0130] As can be seen, by implementing the model verification method based on measured data described in the embodiments of the present invention, the platform performance parameter data is obtained after performing singular value removal, verification, inspection, time alignment, and spatial alignment preprocessing on the measured data of different platforms. This data is then used for the verification and evaluation of the platform simulation model, thereby improving the realism of the simulation model.

[0131] Optionally, in step S21 above, processing the measured dataset using a data elimination model to obtain a first preprocessed dataset includes:

[0132] S211, Obtain the data sample size value of the measured dataset;

[0133] S212, determine whether the data sample size is greater than a first threshold, and obtain a first determination result;

[0134] S213, when the first judgment result is yes, execute S214;

[0135] If the first judgment result is negative, execute S1;

[0136] S214, determine whether the data sample size is less than the second threshold, and obtain the second determination result;

[0137] When the second judgment result is yes, the measured dataset is processed using the first data elimination model to obtain the first preprocessed dataset;

[0138] The expression for the first data removal model is:

[0139] ;

[0140] in, This represents the data sample size of the measured dataset; Represents the first in the measured dataset One measured data point; Indicates the index of the measured data; S represents the mean of the measured dataset; S represents the standard deviation of the measured data.

[0141] It should be noted that when ,but These are singular values ​​and should be discarded.

[0142] When the second judgment result is negative, the measured dataset is processed using the second data elimination model to obtain the first preprocessed dataset;

[0143] The expression for the second data removal model is:

[0144] ;

[0145] in, This represents the data sample size of the measured dataset; Represents the first in the measured dataset One measured data point; Indicates the index of the measured data; This represents the average value of the measured dataset; Indicates the significance level; Indicates the first relation coefficient; This represents the standard deviation of the measured dataset;

[0146] It should be noted that, in this embodiment, the significance level... Set to 0.05;

[0147] The first relationship coefficient ,in accordance with and The critical value table for the Grubbs test was consulted.

[0148] As can be seen, by implementing the model verification method based on measured data described in the embodiments of the present invention, singular values ​​are removed from the measured data of different platforms, providing data support for the subsequent verification and evaluation of platform simulation models, and improving the realism of the simulation models.

[0149] Optionally, in step S22 above, processing the first preprocessed dataset using a data validation model to obtain a second preprocessed dataset includes:

[0150] S221, Obtain the data sample size value of the first preprocessed dataset;

[0151] S222, determine whether the data sample size is greater than the third threshold, and obtain the third judgment result;

[0152] S223, when the third judgment result is yes, execute S224;

[0153] If the result of the third judgment is negative, execute S1;

[0154] S224, determine whether the data sample size is less than the fourth threshold, and obtain the fourth judgment result;

[0155] When the fourth judgment result is yes, the first preprocessed dataset is processed using the first data verification model to obtain the second preprocessed dataset;

[0156] The expression for the first data verification model is:

[0157] ;

[0158] in, This represents the sample order statistics for the first preprocessed dataset; This represents the number of data samples in the first preprocessed dataset; Indicates the first The significance level of the first preprocessed data; Indicates the significance level quantile; Indicates the first relation coefficient;

[0159] It should be noted that the significance level quantiles in accordance with The critical value table for the Grubbs test was consulted.

[0160] It should be noted that the significance level of the first preprocessed data The critical value table for the Grubbs test was consulted.

[0161] When the fourth judgment result is negative, the first preprocessed dataset is processed using the second data verification model to obtain the second preprocessed dataset;

[0162] The expression for the second data verification model is:

[0163]

[0164] in, This represents the sample order statistics for the first preprocessed dataset; This represents the number of data samples in the first preprocessed dataset; Indicates the first The significance level of the first preprocessed data; This indicates the rejection field for the second data verification;

[0165] It should be noted that the rejection field of the second data validation... Depend on The critical value was obtained by consulting the Grubbs test critical value table.

[0166] As can be seen, by implementing the model verification method based on measured data described in the embodiments of the present invention, the measured data from different platforms are verified after outlier removal, resulting in a second preprocessed dataset after verification. This provides data support for the subsequent verification and evaluation of the platform simulation model and improves the realism of the simulation model.

[0167] Optionally, in step S23 above, the step of performing validation processing on the second preprocessed dataset to obtain the third preprocessed dataset includes:

[0168] S231, Obtain the second preprocessed dataset;

[0169] S232, perform autocorrelation processing on each element of the second preprocessed dataset to obtain the correlation dataset;

[0170] The autocorrelation processing expression is:

[0171] ;

[0172] in, The second preprocessed dataset represents the first... The second preprocessed data;

[0173] S233, perform estimation processing on each element of the relevant processing dataset to obtain a set of relevant data estimates;

[0174] The processing expression is:

[0175] ;

[0176] in, It is expressed as the autocorrelation coefficient of white noise; Represents the first in the relevant dataset One relevant data point; This indicates the number of elements in the relevant dataset;

[0177] S234, the estimated value set of the relevant data is judged and processed to obtain the third preprocessed dataset.

[0178] As can be seen, the model verification method based on measured data described in the embodiments of the present invention performs verification processing on the measured data of different platforms after singular value removal and verification processing, providing data support for the subsequent verification and evaluation of platform simulation models and improving the realism of simulation models.

[0179] Optionally, in step S25 above, performing time alignment processing on the fourth preprocessed dataset to obtain the fifth preprocessed dataset includes:

[0180] S251, the fourth preprocessed dataset is parsed to obtain the radar measurement dataset and the preprocessed dataset;

[0181] It should be noted that the parsing process means determining whether any piece of the fourth preprocessed data in the fourth preprocessed dataset is radar measurement data based on the data type. When the fourth preprocessed data is radar measurement data, the fourth preprocessed data is extracted to obtain radar measurement data; all the radar measurement data are removed from the fourth preprocessed dataset to obtain the preprocessed dataset.

[0182] S252, Incrementally sort the radar measurement dataset to obtain a preprocessed radar measurement dataset;

[0183] It should be noted that the incremental sorting refers to sorting by measurement accuracy.

[0184] S253, using a data precision adjustment model, the preprocessed radar measurement data in the preprocessed radar measurement dataset is processed to obtain the radar preprocessed dataset;

[0185] Optionally, the data precision adjustment model expression is:

[0186] ;

[0187] in, Represents a high-precision time index; Indicates a low-precision time index; Indicates the first The coordinates of the x, y, and z directions at time; Indicates the first The velocity in the x, y, and z directions at time 1; Indicates the first Interpolation time; Indicates the first time corresponding to the interpolation time A high-precision time; Indicates the first To the The time difference between moments; for High precision The target coordinates after low-precision alignment, i.e., radar preprocessing data;

[0188] Optionally, the data precision adjustment model expression is:

[0189] ;

[0190] in, express Radar measurement data at any given time; express Radar measurement data at any given time; express Radar measurement data at any given time; express Radar measurement data at any given time, i.e., radar preprocessed data;

[0191] S254, The preprocessed dataset and the radar preprocessed dataset are fused to obtain the fifth preprocessed dataset;

[0192] It should be noted that the fusion process refers to sorting the preprocessed dataset and the radar preprocessed dataset in chronological order.

[0193] As can be seen, the model verification method based on measured data described in the embodiments of the present invention performs time alignment processing on the measured data of different platforms after singular value processing, verification processing and validation processing, which provides data support for the subsequent verification and evaluation of platform simulation models and improves the realism of simulation models.

[0194] Optionally, in step S3 above, processing the first dataset to obtain the second and third datasets includes:

[0195] S31, parse the first dataset to obtain the test parameter dataset and the test result dataset;

[0196] S32, using a sensitivity analysis model, the test parameter dataset and the test result dataset are processed to obtain a second dataset and a third dataset;

[0197] The sensitivity analysis model expression is:

[0198] ;

[0199] ;

[0200] in, Indicates the first mutual information value; Indicates the second mutual information value; This represents the second dataset; This refers to the third dataset; Represents random events for The probability of; Represents random events for The probability of; Represents random events and random events The joint probability.

[0201] As can be seen, by implementing the model verification method based on measured data described in the embodiments of the present invention, the first dataset after preprocessing is processed to obtain the second dataset and the third dataset, which provides data support for the subsequent verification and evaluation of the platform simulation model and improves the realism of the simulation model.

[0202] Optionally, in step S4 above, processing the second dataset to obtain the fourth dataset includes:

[0203] S41, process the second dataset to obtain the second optimized dataset;

[0204] S42, using the parameter control model, iteratively process the second optimized dataset to obtain the fourth dataset;

[0205] The expression for the parameter control model is:

[0206] ;

[0207] ;

[0208] in, Indicates the first Parameters at the next iteration The value; Indicates the first Parameters at the next iteration The value; Indicates the learning rate; Represents the prediction function; This indicates that the prediction function has respect to the parameters. In the Partial derivative at the next iteration; Indicates the first Secondary parameter update step size; Represents the loss function; Represents the loss function Regarding parameters In the The gradient at the next iteration;

[0209] It should be noted that, in this embodiment, the number of iterations is set to 200; the learning rate... Set to 0.05;

[0210] As can be seen, by implementing the model verification method based on measured data described in the embodiments of the present invention, the second dataset is processed to obtain the fourth dataset, which provides data support for the subsequent verification and evaluation of the platform simulation model and improves the realism of the simulation model.

[0211] Optionally, in step S5 above, processing the third and fourth datasets to obtain model evaluation result information includes:

[0212] S51, Obtain the target model;

[0213] It should be noted that the target model refers to the user model used for evaluation;

[0214] It should be noted that the model evaluation results are used to evaluate simulation tests. By constructing a model fidelity evaluation index system and importing the model fidelity evaluation algorithm into the system, the system performs comparative analysis of model test data and measured data for radar, communication, radar jamming, and communication jamming models. The final evaluation results are displayed in text, tables, and graphs. The optimized parameter set needs to be re-integrated into the simulation system, and simulation experiments should be conducted under given combat conditions to reconfirm consistency with measured data. By comparing the consistency between the simulation model experiment and the field test output under the same input conditions and operating environment, the reliability and usability of the model are evaluated. Finally, the final state of the model parameters is confirmed and solidified, resulting in a model that closely resembles the actual equipment.

[0215] S52, the third dataset is input into the target model, and the simulation result dataset is obtained after processing by the target model;

[0216] Optionally, the confidence interval method can be used for model evaluation, specifically including:

[0217] Step 1: Based on the interval Set decision criteria for the simulation output results, including frequency thresholds. and the confidence level of parameter estimation ;

[0218] Step 2: Calculate the confidence intervals for all simulation output samples. And satisfy ;

[0219] Step 3: Event Statistics probability of occurrence The simulation results dataset is obtained;

[0220] Optionally, fuzzy comprehensive evaluation method can be used for model evaluation, specifically including:

[0221] 1) Establish a set of factors for the evaluation object The evaluation factors are the various attributes or performance of the object. In different contexts, they are also called parameter indicators or quality indicators. They can comprehensively reflect the quality of the object, and therefore the object can be evaluated by these factors.

[0222] It should be noted that, A positive integer, representing the number of factors in the evaluation object;

[0223] 2) Establish an evaluation result set ;

[0224] It should be noted that the evaluation result set includes The evaluation results of the factors of the evaluation objects; A positive integer, representing the number of evaluation results;

[0225] It should be noted that the evaluation result set is a set of evaluation levels, representing a set of fitness levels;

[0226] For example, the evaluation result represents the training result as excellent, good, average, or poor;

[0227] 3) Using a fuzzy mapping model, the evaluation result set and the set of evaluation object factors are processed to obtain the factor evaluation matrix:

[0228] The expression for the fuzzy mapping model is:

[0229]

[0230]

[0231] ;

[0232] in, Represents a single-factor evaluation matrix; Indicates the number of factors in the evaluation object; This indicates the number of evaluation results for the factors of the evaluated object; An index representing the factors being evaluated; An index representing the evaluation results; Indicates the first The first factor of the evaluation object The fitness value of the assessment results; Indicates the first Each evaluation object is a factor; This represents the set of factors that will be evaluated.

[0233] 4) Assign different weights to each of the evaluation factors to obtain a factor weight set. ;

[0234] in, Indicates the first The weight values ​​of each evaluation factor; This indicates the number of weights for the factors; An index representing the weight of the factor;

[0235] 5) Process the factor evaluation matrix and the factor weight set using a comprehensive evaluation model to obtain the simulation result dataset;

[0236] The expression for the comprehensive evaluation model is:

[0237] ;

[0238] in, This represents a fuzzy subset of the evaluation result set; Represents a fuzzy composition operator;

[0239] S53, evaluate the simulation result dataset and the fourth dataset to obtain model evaluation result information;

[0240] It should be noted that when the probability of an event occurring is greater than a set threshold, the result of the simulation experiment is accepted; otherwise, the result of the simulation experiment is not accepted.

[0241] or The largest element in the matrix represents the model evaluation result information;

[0242] It should be noted that the evaluation process refers to comparing the consistency between the simulation result dataset and the fourth dataset, and evaluating the credibility and usefulness of the target model;

[0243] It should be noted that at the end of the system performance evaluation task, the evaluation results of the model are presented intuitively in various forms such as text, tables, curves, and graphs.

[0244] As can be seen, by implementing the model verification method based on measured data described in the embodiments of the present invention, processing the third and fourth datasets, and obtaining model evaluation results, the realism of the simulation model is improved.

[0245] Example 2

[0246] Please see Figure 3 , Figure 3This is a schematic diagram of a model verification device based on measured data disclosed in an embodiment of the present invention. Figure 3 The described apparatus can be applied to simulation systems, such as local servers or cloud servers used in simulation systems, and the embodiments of the present invention are not limited thereto. Figure 3 As shown, the device may include:

[0247] Data acquisition module 101 is used to acquire the measured dataset;

[0248] The first processing module 102 is used to process the measured dataset to obtain a first dataset;

[0249] The second processing module 103 is used to process the first dataset to obtain a second dataset and a third dataset;

[0250] The third processing module 104 is used to process the second dataset to obtain the fourth dataset;

[0251] The fourth processing module 105 is used to process the third dataset and the fourth dataset to obtain model evaluation result information.

[0252] Example 3

[0253] Please see Figure 4 , Figure 4 This is a schematic diagram of a model verification device based on measured data disclosed in an embodiment of the present invention. Figure 4 The described device can be applied to simulation systems, such as local servers or cloud servers used in simulation systems, and the embodiments of the present invention are not limited thereto. Figure 4 As shown, the device may include:

[0254] Memory 201 storing executable program code;

[0255] Processor 202 coupled to memory 201;

[0256] The processor 202 calls the executable program code stored in the memory 201 to execute the steps in the model verification method based on measured data described in Embodiment 1.

[0257] Example 4

[0258] This invention discloses a computer-readable storage medium that stores a computer program for electronic data interchange, wherein the computer program causes a computer to execute the steps in the model verification method based on measured data described in Embodiment 1.

[0259] Example 5

[0260] This invention discloses a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to cause a computer to perform the steps in the model verification method based on measured data described in Embodiment 1.

[0261] The device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules; that is, they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0262] Through the detailed description of the above embodiments, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus necessary general-purpose hardware platform, or of course by hardware. Based on this understanding, the above-mentioned technical solution, or the part that contributes to the existing technology, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, including read-only memory (ROM), random access memory (RAM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), one-time programmable read-only memory (OTPROM), electrically erasable programmable read-only memory (EEPROM), compact-disc read-only memory (CD-ROM) or other optical disc storage, disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0263] Finally, it should be noted that the model verification method, system, and apparatus based on measured data disclosed in the embodiments of the present invention are merely preferred embodiments of the present invention and are only used to illustrate the technical solutions of the present invention, not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A model verification method based on measured data, characterized in that, The method includes: S1, Obtain the actual test dataset; The measured dataset, which consists of measured data collected from different platforms, is used to verify the radar simulation model. S2, process the measured dataset to obtain the first dataset; S3, process the first dataset to obtain the second dataset and the third dataset; S4, process the second dataset to obtain the fourth dataset; S5, process the third and fourth datasets to obtain model evaluation results; The step of processing the measured dataset to obtain the first dataset includes: S21, The measured dataset is processed using a data elimination model to obtain a first preprocessed dataset; The data removal model includes a first data removal model and a second data removal model; S22, using a data validation model, the first preprocessed dataset is processed to obtain a second preprocessed dataset; The data verification model includes a first data verification model and a second data verification model; S23, perform validation processing on the second preprocessed dataset to obtain the third preprocessed dataset; S24, perform error processing on the third preprocessed dataset to obtain the fourth preprocessed dataset; S25, perform time alignment processing on the fourth preprocessed dataset to obtain the fifth preprocessed dataset; S26, Spatial alignment processing is performed on the fifth preprocessed dataset to obtain the first dataset; The step of using a data removal model to process the measured dataset to obtain a first preprocessed dataset includes: S211, Obtain the data sample size value of the measured dataset; wherein, the data sample is a remote sensing image; S212, determine whether the data sample size is greater than a first threshold, and obtain a first determination result; S213, when the first judgment result is yes, execute S214; If the first judgment result is negative, execute S1; S214, determine whether the data sample size is less than the second threshold, and obtain the second determination result; When the second judgment result is yes, the measured dataset is processed using the first data elimination model to obtain the first preprocessed dataset; The expression for the first data removal model is: ； in, This represents the data sample size of the measured dataset; Represents the first in the measured dataset One measured data point; Indicates the index of the measured data; S represents the mean of the measured dataset; S represents the standard deviation of the measured data. When the second judgment result is negative, the measured dataset is processed using the second data elimination model to obtain the first preprocessed dataset; The expression for the second data removal model is: ； in, Indicates the significance level; Indicates the first relation coefficient; This represents the standard deviation of the measured dataset; The step of processing the first preprocessed dataset using a data validation model to obtain the second preprocessed dataset includes: S221, Obtain the data sample size value of the first preprocessed dataset; S222, determine whether the data sample size is greater than the third threshold, and obtain the third judgment result; S223, when the third judgment result is yes, execute S224; If the result of the third judgment is negative, execute S1; S224, determine whether the data sample size is less than the fourth threshold, and obtain the fourth judgment result; When the fourth judgment result is yes, the first preprocessed dataset is processed using the first data verification model to obtain the second preprocessed dataset; The expression for the first data verification model is: ； in, This represents the sample order statistics for the first preprocessed dataset; Indicates the first The significance level of the first preprocessed data; Indicates the significance level quantile; When the fourth judgment result is negative, the first preprocessed dataset is processed using the second data verification model to obtain the second preprocessed dataset; The expression for the second data verification model is: ； in, This indicates the rejection field for the second data verification; The fourth preprocessed dataset is subjected to time alignment processing to obtain the fifth preprocessed dataset, which includes: S251, the fourth preprocessed dataset is parsed to obtain the radar measurement dataset and the preprocessed dataset; S252, Incrementally sort the radar measurement dataset to obtain a preprocessed radar measurement dataset; S253, using a data precision adjustment model, the preprocessed radar measurement data in the preprocessed radar measurement dataset is processed to obtain the radar preprocessed dataset; S254, The preprocessed dataset and the radar preprocessed dataset are fused to obtain the fifth preprocessed dataset; The process of processing the first dataset to obtain the second and third datasets includes: S31, parse the first dataset to obtain the test parameter dataset and the test result dataset; S32, using a sensitivity analysis model, the test parameter dataset and the test result dataset are processed to obtain a second dataset and a third dataset; The sensitivity analysis model expression is: ; ; in, Indicates the first mutual information value; Indicates the second mutual information value; This represents the second dataset; This refers to the third dataset; Table of random events for The probability of; Represents random events for The probability of; Represents random events and random events The joint probability; The process of processing the second dataset to obtain the fourth dataset includes: S41, process the second dataset to obtain the second optimized dataset; S42, using the parameter control model, iteratively process the second optimized dataset to obtain the fourth dataset; The expression for the parameter control model is: ; ; in, Indicates the first Parameters at the next iteration The value; Indicates the first Parameters at the next iteration The value; Indicates the learning rate; Represents the prediction function; This indicates that the prediction function has respect to the parameters. In the Partial derivative at the next iteration; Indicates the first Secondary parameter update step size; Represents the loss function; Represents the loss function Regarding parameters In the The gradient at the next iteration.

2. The model verification method based on measured data according to claim 1, characterized in that, The step of performing validation processing on the second preprocessed dataset to obtain the third preprocessed dataset includes: S231, perform autocorrelation processing on each element of the second preprocessed dataset to obtain the correlation dataset; The autocorrelation processing expression is: ； S232, perform estimation processing on each element of the relevant processing dataset to obtain a set of relevant data estimates; The estimation processing expression is: ； in, It is expressed as the autocorrelation coefficient of white noise; S233, the estimated value set of the relevant data is judged and processed to obtain the third preprocessed dataset.

3. A model verification device based on measured data, used to execute the model verification method based on measured data as described in claim 1, characterized in that, The device includes: The data acquisition module is used to acquire the actual test dataset; The first processing module is used to process the measured dataset to obtain the first dataset; The second processing module is used to process the first dataset to obtain the second dataset and the third dataset; The third processing module is used to process the second dataset to obtain the fourth dataset; The fourth processing module is used to process the third and fourth datasets to obtain model evaluation results.

4. A model verification device based on measured data, characterized in that, The device includes: Memory containing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the model verification method based on measured data as described in any one of claims 1-2.

5. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions, which, when invoked, are used to execute the model verification method based on measured data as described in any one of claims 1-2.

Images