Physical entity operation state prediction method and device based on digital twin model, equipment, medium and product

Abstract

The invention discloses a physical entity operation state prediction method and device based on a digital twinning model, equipment, a medium and a product, and relates to the technical field of digital twinning, and the method comprises the steps: building the digital twinning model based on the technical state data of a physical entity; performing mapping training on the digital twinborn model through the historical technical state data and the historical operation state data of the physical entity to obtain a trained digital twinborn model; acquiring real-time technical state data of the physical entity, and determining predicted operation state data of the physical entity based on the trained digital twin model; performing iterative correction on the trained digital twinborn model based on the real-time technical state data and the predicted operation state data; and predicting the operation state of the physical entity based on the corrected digital twin model. According to the method, the digital twin model can be corrected in real time, so that the digital twin model is consistent with the operation state of the physical entity, and the prediction precision of the operation state of the physical entity is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of digital twinning, in particular to a physical entity running state prediction method and device based on a digital twinning model, equipment, medium and product. BACKGROUND

[0002] As a technology that closely combines physical entities and virtual models, digital twinning has been widely applied in many fields such as industrial manufacturing, urban planning, intelligent transportation, etc. The digital twinning model needs to reflect the state changes of the physical entity in real time. However, in the actual operation process, due to factors such as sensor errors, data transmission delays, model simplification and differences between the actual physical process, the state of the digital twinning model and the physical entity often does not match, which seriously affects the reliability and application effect of the digital twinning technology, thereby affecting the accuracy of the physical entity running state prediction. SUMMARY

[0003] The purpose of the present application is to provide a physical entity running state prediction method, device, equipment, medium and product based on a digital twinning model, which can correct the digital twinning model in real time, so that the running state of the digital twinning model and the physical entity is consistent, thereby improving the accuracy of the physical entity running state prediction.

[0004] To achieve the above purpose, the present application provides the following solutions:

[0005] In a first aspect, the present application provides a physical entity running state prediction method based on a digital twinning model, comprising:

[0006] constructing a digital twinning model based on technical state data of the physical entity;

[0007] mapping training the digital twinning model through historical technical state data and historical running state data of the physical entity to obtain a trained digital twinning model;

[0008] obtaining real-time technical state data of the physical entity, and determining predicted running state data of the physical entity based on the trained digital twinning model;

[0009] iteratively correcting the trained digital twinning model based on real-time running state data and predicted running state data of the physical entity;

[0010] predicting the running state of the physical entity based on the corrected digital twinning model.

[0011] In a second aspect, the present application provides a physical entity running state prediction device based on a digital twinning model, comprising:

[0012] The construction module is configured to construct a digital twin model based on the technical state data of the physical entity.

[0013] The training module is configured to perform mapping training on the digital twin model by using the historical technical state data and the historical running state data of the physical entity, to obtain a trained digital twin model.

[0014] The first prediction module is configured to obtain real-time technical state data of the physical entity, and determine predicted running state data of the physical entity based on the trained digital twin model.

[0015] The correction module is configured to perform iterative correction on the trained digital twin model based on the real-time running state data and the predicted running state data of the physical entity.

[0016] The second prediction module is configured to predict the running state of the physical entity based on the corrected digital twin model.

[0017] In a third aspect, the present application provides a computer device, comprising a memory, a processor, a computer program stored in the memory and executable on the processor, and the processor executes the computer program to implement the steps of the physical entity running state prediction method based on the digital twin model according to any one of the above.

[0018] In a fourth aspect, the present application provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the steps of the physical entity running state prediction method based on the digital twin model according to any one of the above.

[0019] In a fifth aspect, the present application provides a computer program product, comprising a computer program, and the computer program is executed by a processor to implement the steps of the physical entity running state prediction method based on the digital twin model according to any one of the above.

[0020] According to the embodiments provided in the present application, the following technical effects are disclosed.

[0021] The present application can perform iterative correction on the digital twin model according to the predicted running state data of the digital twin model and the real-time running state data of the physical entity, realize real-time dynamic correction of the digital twin model, ensure synchronization between the digital twin model and the physical entity, and make the digital twin model accurately predict the running state data of the physical entity. BRIEF DESCRIPTION OF DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments. Obviously, the drawings described below only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained from these drawings without creative labor.

[0023] Figure 1 An application environment diagram of a physical entity operation state prediction method based on a digital twin model according to an embodiment of the present application;

[0024] Figure 2 A flowchart of a physical entity operation state prediction method based on a digital twin model according to an embodiment of the present application;

[0025] Figure 3 A functional module diagram of a physical entity operation state prediction device based on a digital twin model according to another embodiment of the present application;

[0026] Figure 4 A structural diagram of a computer device according to an embodiment of the present application. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments only constitute some embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0028] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments.

[0029] The physical entity operation state prediction method based on a digital twin model provided by the embodiments of the present application can be applied to, for example Figure 1The application environment shown. Among them, the terminal 102 communicates with the server 104 through the network. The data storage system can store the data required by the server 104 to process. The data storage system can be set up separately, or integrated on the server 104, or placed on the cloud or other servers. The terminal 102 can send the real-time technical state data of the physical entity to the server 104, and the server 104 receives the real-time technical state data of the physical entity. The server 104 obtains the real-time technical state data of the physical entity, and determines the predicted running state data of the physical entity based on the trained digital twin model; Based on the real-time running state data and the predicted running state data of the physical entity, the trained digital twin model is iteratively corrected; Based on the corrected digital twin model, the running state of the physical entity is predicted. The server 104 can feed back the obtained running state of the physical entity to the terminal 102. In addition, in some embodiments, the physical entity running state prediction method based on the digital twin model can also be implemented by the server 104 or the terminal 102 alone, such as the terminal 102 can directly predict the running state data for the real-time technical state data of the physical entity to be processed, or the server 104 can obtain the real-time technical state data of the physical entity from the data storage system for processing.

[0030] Among them, the terminal 102 can be but not limited to various desktop computers, notebook computers, smart phones, tablet computers, Internet of Things devices and portable wearable devices. The Internet of Things device can be a smart speaker, a smart television, a smart air conditioner, a smart vehicle-mounted device, etc. The portable wearable device can be a smart watch, a smart bracelet, a head-mounted device, etc. The server 104 can be implemented by an independent server or a server cluster composed of multiple servers, and can also be a cloud server.

[0031] In an exemplary embodiment, as Figure 2 shown, a physical entity running state prediction method based on a digital twin model is provided, which is executed by a computer device, specifically by a terminal or a server computer device alone, or by a terminal and a server together. In the embodiment of the present application, the method is applied to the server 104 in Figure 1 The following steps 201 to 205 are described. Among them:

[0032] Step 201, constructing a digital twin model based on technical state data of a physical entity.

[0033] Firstly, technical state data of the physical entity needs to be collected, wherein the technical state data includes topographic posture feature data, behavior action feature data and running parameter data, the topographic posture feature data refers to quantitative data of geometric structure, spatial position and morphological change of the physical entity, such as size, shape, assembly structure, deformation and the like of the equipment; the behavior action feature data refers to time sequence data of behavior logic and process of action performed by the physical entity, such as instruction sequence, working period, time sequence record, acceleration curve and the like; the running parameter data refers to real-time physical quantities such as current, rotating speed, temperature, pressure, vibration and the like collected by sensors in the running process of the physical entity. The digital twin model is constructed by using geometric modeling, physical modeling and data modeling, the geometric modeling is to establish an appearance model of the entity by using 3D laser scanning or CAD data; the physical modeling is to establish a mechanism model based on first principle; and the data modeling is to use a machine learning model (LSTM, random forest and the like) for processing nonlinear relationship and predicting performance degradation.

[0034] In step 202, the digital twin model is mapped and trained by using historical technical state data and historical running state data of the physical entity, to obtain a trained digital twin model.

[0035] The running state data includes accuracy data, stability data and response data.

[0036] The accuracy data is used to measure the closeness of the output result of the physical entity to the true value, such as error, precision and the like, and reflects the correctness and reliability of the result; the stability data is used to measure the ability of the physical entity to keep consistent performance under long-time running or external interference, such as output fluctuation range, fault interval time and the like, and reflects the anti-interference ability and reliability; and the response data is used to measure the real-time processing ability of the physical entity to input signal or instruction, such as response time, data update frequency and the like, and reflects the real-time performance and efficiency.

[0037] In step 203, real-time technical state data of the physical entity is acquired, and predicted running state data of the physical entity is determined based on the trained digital twin model.

[0038] In step 204, the trained digital twin model is iteratively corrected based on real-time running state data and predicted running state data of the physical entity.

[0039] In step 205, the running state of the physical entity is predicted based on the corrected digital twin model.

[0040] By implementing the above steps 201 to 205, the trained digital twin model can be corrected in real time, and the running state of the digital twin model and the physical entity is ensured to be synchronized.

[0041] In one exemplary embodiment, step 201 specifically includes:

[0042] The initial technical state data of the physical entity includes initial topographic posture feature data, initial behavior action feature data, and initial running parameter data, and the initial running state data includes initial accuracy data, initial stability data, and initial response data.

[0043] The digital twin model is constructed according to the initial topographic posture feature data, the initial behavior action feature data, the initial running parameter data, and the initial accuracy data, the initial stability data, and the initial response data. Specifically, the digital twin model adopts an LSTM (Long Short-Term Memory) neural network.

[0044] In an exemplary embodiment, the mapping training of the digital twin model in step 202 specifically includes:

[0045] The historical technical state data and the historical running state data of the physical entity at multiple historical nodes are obtained.

[0046] The historical technical state data and the historical running state data are input into the initial digital twin model for mapping training. In the mapping training process, the control strategy of the digital twin model is optimized through reinforcement learning. Specifically, the historical technical state data is input into the initial digital twin model, the error of the digital twin model is dynamically corrected through a parameter identification algorithm (such as particle swarm optimization), the state of the digital twin model is synchronized with that of the physical entity, and thus a trained digital twin model is obtained.

[0047] In an exemplary embodiment, step 204 specifically includes steps 301-302:

[0048] Step 301: Determine the evaluation state according to the real-time running state data and the predicted running state data. Specifically, steps 401-405 are included:

[0049] Step 401: Determine the weight coefficient of the real-time running state data according to the real-time running state data.

[0050] The weight coefficient of the real-time running state data is determined by querying a preset weight coefficient list according to the real-time running state data. The weight coefficient specifically includes an accuracy weight coefficient, a stability weight coefficient, and a response weight coefficient.

[0051] The preset weight coefficient list is obtained by querying a third-party preset data virtual processing platform. The third-party preset data virtual processing platform is a platform for virtually processing preset data, which usually combines virtualization, data processing, and analysis technologies to achieve efficient processing and analysis of data, such as LabVIEW.

[0052] Step 402, determining data deviation rates according to real-time running state data and predicted running state data respectively.

[0053] The real-time running state data and the predicted running state data are compared one by one to obtain accuracy deviation rates, stability deviation rates and response deviation rates.

[0054] The accuracy deviation rate is the ratio of the absolute value of the difference between the real-time accuracy data and the predicted accuracy data to the real-time accuracy data; the stability deviation rate is the ratio of the absolute value of the difference between the real-time stability data and the predicted stability data to the real-time stability data; and the response deviation rate is the ratio of the absolute value of the difference between the real-time response data and the predicted response data to the real-time response data.

[0055] Step 403, determining an evaluation index according to the weight coefficient and the data deviation rate.

[0056] The evaluation index calculation formula is:

[0057] z y =α1z q x z +α2w h x w +α3y x x y .

[0058] Wherein, z y is the evaluation index, z q , w h , y x are the accuracy deviation rate, the stability deviation rate and the response deviation rate respectively, x z , x w , x y are the accuracy weight coefficient, the stability weight coefficient and the response weight coefficient respectively, and α1, α2, α3 are preset characteristic coefficients, which are obtained by querying a third-party preset data virtual processing platform.

[0059] Step 404, determining a state consistency relative value according to the evaluation index and a preset state consistency reference index.

[0060] Specifically, the evaluation index is compared with the preset state consistency reference index to obtain the state consistency relative value; for example, the obtained evaluation index is 4, and the preset state consistency reference index is 5, then 4 / 5=0.8, and the state consistency relative value is 0.8.

[0061] Step 405, determining the evaluation state according to the state consistency relative value and the preset state consistency evaluation threshold. Specifically, the state consistency relative value is compared with the preset state consistency evaluation threshold to obtain the evaluation state, which includes state consistency or state inconsistency.

[0062] When the state consistency relative value is less than or equal to the preset state consistency evaluation threshold, the evaluation state is determined to be state inconsistency.

[0063] When the state consistency relative value is greater than the preset state consistency evaluation threshold, the evaluation state is determined to be state consistency.

[0064] In this embodiment, the preset state consistency evaluation threshold is set to 0.85. When the state consistency relative value belongs to the range (0, 0.85], the evaluation state is state inconsistency; when the state consistency relative value belongs to the range (0.85, 1], the evaluation state is state consistency. For example, the obtained state consistency relative value is 0.8, which is less than the preset state consistency evaluation threshold 0.85, so the evaluation state is determined to be state inconsistency. If the obtained state consistency relative value is 0.9, which is greater than the preset state consistency evaluation threshold, the evaluation state is determined to be state consistency.

[0065] Step 302, when the evaluation state is state inconsistency, the trained digital twin model is iteratively corrected using the gradient descent algorithm.

[0066] 1. Set the bias minimization objective function

[0067] Let the expression of the trained digital twin model be M(θ), where θ=[θ1,θ2,......,θ n ] is the key parameter to be corrected, and n is the number of key parameters.

[0068] The target function is constructed by taking the minimum bias between the actual value of the running state data and the predicted value of the running state data by the trained digital twin model as the target, and the expression of the target function is:

[0069]

[0070] Where x i , y i are the actual values of the technical state data and the running state data, respectively, N is the number of samples, and L(θ) is the target function.

[0071] 2. Gradient calculation and optimization strategy

[0072] (1) Gradient calculation: calculate the gradient of the target function with respect to the key parameters:

[0073]

[0074] (2) Gradient optimization strategy:

[0075] A. Introduce momentum term to reduce oscillation:

[0076]

[0077] θ t+1 = θ t - v t+1 ;

[0078] where γ is the momentum coefficient, η is the learning rate, v t+1 is the momentum at t+1, g t is the gradient of the key parameter θ t+1 at t, and θ t is the key parameter at t+1.

[0079] B. Adaptive learning rate adjustment:

[0080]

[0081] where EMA is the exponential moving average, which solves the parameter scale difference problem, and ∈ is a constant.

[0082] 3. Update the key parameters according to the calculated gradient and perform iterative calculation until the iteration terminates.

[0083] Key parameter update rule:

[0084]

[0085] The learning rate η can be dynamically adjusted by the following method:

[0086]

[0087] where η t is the learning rate at t, and η0 is the initial learning rate.

[0088] The following three convergence criteria must be met simultaneously to terminate the iteration.

[0089] (1) Loss change threshold:

[0090] |L(θ t )-L(θ t-1 )|<∈ L , ∈ L =10 -6 , for K consecutive iterations.

[0091] (2) Gradient norm threshold:

[0092]

[0093] (3) maximum number of iterations:

[0094] Prevent infinite loop, set T max 10000 times.

[0095] 4. After the iteration is terminated, the parameter value of the key parameter obtained is the key parameter of the digital twin model after being corrected by the gradient descent algorithm.

[0096] In an exemplary embodiment, after step 204, steps 501-502 are further included:

[0097] Step 501: Determine the updated evaluation state according to the updated real-time running state data and the corresponding predicted running state data.

[0098] Obtain the real-time technical state data of the preset time after correction, and obtain the twin iterative running state data, i.e., the corresponding predicted running state data, by using the corrected digital twin model; the twin iterative running state data includes iterative accuracy data, iterative stability data, and iterative response data.

[0099] According to the iterative accuracy data, the iterative stability data, and the iterative response data, the real-time accuracy data, the real-time stability data, and the real-time response data, and the accuracy weight coefficient, the stability weight coefficient, and the response weight coefficient, process to obtain an iterative consistency evaluation index, and determine the updated evaluation state according to the iterative consistency evaluation index.

[0100] The iterative consistency evaluation index calculation formula is:

[0101]

[0102] wherein, d z is the iterative consistency evaluation index, z d , w d , x d are the iterative accuracy data, the iterative stability data, and the iterative response data, respectively, z s , w s , x s are the real-time accuracy data, the real-time stability data, and the real-time response data, respectively, x z , x w , x y are the accuracy weight coefficient, the stability weight coefficient, and the response weight coefficient, respectively, and β1, β2, β3 are preset characteristic coefficients.

[0103] The iterative consistency evaluation index is compared with the preset state consistency benchmark index to obtain an iterative consistency relative value.

[0104] The iterative consistency relative value is compared with a preset state consistency evaluation threshold value in a threshold comparison to obtain an updated evaluation state, i.e., an iterative correction effectiveness, including iterative correction effectiveness or iterative correction ineffectiveness.

[0105] If the iterative consistency relative value is less than or equal to the preset state consistency evaluation threshold value, it is determined that the iterative correction effectiveness is iterative correction ineffectiveness.

[0106] If the iterative consistency relative value is greater than the preset state consistency evaluation threshold value, it is determined that the iterative correction effectiveness is iterative correction effectiveness.

[0107] At step 502, the effectiveness of the iterative correction of the corrected digital twin model is determined according to the updated evaluation state. Specifically, steps 601-604 are included:

[0108] At step 601, a number of times of continuous iterative correction ineffectiveness in a preset time period after the iterative correction is obtained.

[0109] At step 602, the number of times is compared with a preset threshold number of times.

[0110] At step 603, if the number of times is less than or equal to the preset threshold number of times, the iterative correction of the trained digital twin model is maintained, i.e., the iterative correction of the corrected digital twin model is effective.

[0111] At step 604, if the number of times is greater than the preset threshold number of times, the iterative correction of the trained digital twin model is stopped, i.e., the iterative correction of the corrected digital twin model is ineffective and an early warning is output.

[0112] For example, after the iterative correction, the effectiveness of the iterative correction is determined. If iterative correction ineffectiveness occurs continuously for multiple times, a maintenance personnel needs to manually maintain, the number of times of continuous iterative correction ineffectiveness in a preset time period is obtained, and the number of times is compared with a preset threshold number of times. In this embodiment, the preset time period is set to 15 minutes, and the preset threshold number of times is set to 3 times. For example, the obtained number of times is 2 times, which is less than the preset threshold number of times. Therefore, the iterative correction of the digital twin model is maintained. If the obtained number of times is 4 times, which is greater than the preset threshold number of times, it indicates that the digital twin model has a serious problem. Therefore, the iterative correction of the digital twin model is stopped, and an early warning is output.

[0113] Based on the same inventive concept, the embodiments of the present application also provide a digital-twin-model-based physical entity operation state prediction apparatus for implementing the above-mentioned digital-twin-model-based physical entity operation state prediction method. The solution provided by the apparatus is similar to the implementation solution described in the above-mentioned method, and therefore the specific limitations in one or more digital-twin-model-based physical entity operation state prediction apparatus embodiments provided below can refer to the limitations of the method described above, and will not be described here again.

[0114] In one exemplary embodiment, as shown in Figure 3 A digital-twin-model-based physical entity operation state prediction apparatus is provided, including:

[0115] A construction module 31 is configured to construct a digital twin model based on technical state data of a physical entity.

[0116] A training module 32 is configured to perform mapping training on the digital twin model based on historical technical state data and historical operation state data of the physical entity, to obtain a trained digital twin model.

[0117] A first prediction module 33 is configured to obtain real-time technical state data of the physical entity, and determine predicted operation state data of the physical entity based on the trained digital twin model.

[0118] A correction module 34 is configured to perform iterative correction on the trained digital twin model based on real-time operation state data and predicted operation state data of the physical entity.

[0119] A second prediction module 35 is configured to predict the operation state of the physical entity based on the corrected digital twin model.

[0120] In one exemplary embodiment, a computer device is provided, which can be a server or a terminal, and an internal structure diagram of the computer device can be as shown in Figure 4As shown in the figure. The computer device includes a processor, a memory, an input / output interface (I / O for short) and a communication interface. Among them, the processor, the memory and the input / output interface are connected through the system bus, and the communication interface is connected to the system bus through the input / output interface. Among them, the processor of the computer device is used to provide computing and control capability. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store real-time technical state data of the physical entity. The input / output interface of the computer device is used to exchange information between the processor and external devices. The communication interface of the computer device is used to communicate with the terminal outside through the network connection. The computer program is executed by the processor to realize the physical entity running state prediction method based on the digital twin model.

[0121] Those skilled in the art can understand that, Figure 4 The structure shown in the figure is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different component arrangement.

[0122] In one exemplary embodiment, a computer device is also provided, including a memory and a processor, the memory storing a computer program, and the processor executing the computer program to realize the steps in the above method embodiments.

[0123] In one exemplary embodiment, a computer readable storage medium is provided, storing a computer program, which is executed by a processor to realize the steps in the above method embodiments.

[0124] In one exemplary embodiment, a computer program product is provided, including a computer program, which is executed by a processor to realize the steps in the above method embodiments.

[0125] It should be noted that the user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or authorized by all parties, and the collection, use and processing of related data need to comply with relevant regulations.

[0126] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical storage, high-density embedded non-volatile memory, resistive memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration but not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0127] The database involved in the embodiments provided in the present application can include at least one of a relational database and a non-relational database. The non-relational database can include a distributed database based on a blockchain, etc., without being limited thereto. The processor involved in the embodiments provided in the present application can be a general-purpose processor, a central processing unit, a graphics processing unit, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., without being limited thereto.

[0128] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combinations of the technical features do not exist contradictory, they should be considered as the scope of the present application.

[0129] The principles and implementation modes of the present application are described by applying specific examples herein, and the above-mentioned embodiments are only used to help understand the method and its core idea of the present application; meanwhile, for those skilled in the art, according to the idea of the present application, the specific implementation mode and application range can be changed. In conclusion, the content of the present application should not be understood as a limitation.

Claims

1. A method for predicting the operating state of a physical entity based on a digital twin model, characterized in that, The method comprises the following steps: constructing a digital twin model based on technical state data of a physical entity; training the digital twin model based on historical technical state data and historical running state data of the physical entity, to obtain a trained digital twin model; obtaining real-time technical state data of the physical entity, and determining predicted running state data of the physical entity based on the trained digital twin model; iteratively correcting the trained digital twin model based on real-time running state data and predicted running state data of the physical entity; predicting the running state of the physical entity based on the corrected digital twin model. 2.The physical entity operation state prediction method based on digital twin model according to claim 1, characterized in that, The technical state data comprises topographic attitude feature data, behavior action feature data and running parameter data, and the running state data comprises accuracy data, stability data and response data. 3.The physical entity operation state prediction method based on digital twin model according to claim 1, wherein, The iteratively correcting the trained digital twin model based on real-time running state data and predicted running state data of the physical entity comprises the following steps: determining an evaluation state according to the real-time running state data and the predicted running state data; when the evaluation state is inconsistent, iteratively correcting the trained digital twin model by using a gradient descent algorithm.

4. The method of claim 3, wherein the method further comprises: The determining an evaluation state according to the real-time running state data and the predicted running state data comprises the following steps: determining a weight coefficient of the real-time running state data according to the real-time running state data; determining a data deviation rate according to the real-time running state data and the predicted running state data respectively; determining an evaluation index according to the weight coefficient and the data deviation rate; determining a state consistency relative value according to the evaluation index and a preset state consistency benchmark index; determining the evaluation state according to the state consistency relative value and a preset state consistency evaluation threshold.

5. The method of claim 4, wherein the method further comprises: The determining the evaluation state according to the state consistency relative value and a preset state consistency evaluation threshold comprises the following steps: when the state consistency relative value is less than or equal to the preset state consistency evaluation threshold, determining that the evaluation state is inconsistent; when the state consistency relative value is greater than the preset state consistency evaluation threshold, determining that the evaluation state is consistent.

6. The method of claim 1, wherein the method further comprises: After iteratively correcting the trained digital twin model based on real-time running state data and predicted running state data of the physical entity, the method further comprises the following steps: determining an updated evaluation state according to updated real-time running state data and corresponding predicted running state data; determining the effectiveness of the iteratively correcting the digital twin model according to the updated evaluation state. 7.A device for predicting a running state of a physical entity based on a digital twin model, characterized in that, The method comprises the following steps: a constructing module, configured to construct a digital twin model based on technical state data of a physical entity; a training module, configured to train the digital twin model based on historical technical state data and historical running state data of the physical entity, to obtain a trained digital twin model; a first predicting module, configured to obtain real-time technical state data of the physical entity, and determine predicted running state data of the physical entity based on the trained digital twin model; a correcting module, configured to iteratively correct the trained digital twin model based on real-time running state data and predicted running state data of the physical entity; a second predicting module, configured to predict the running state of the physical entity based on the corrected digital twin model.

8. A computer device comprising: A memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that the processor executes the computer program to implement the method for predicting the running state of a physical entity based on a digital twin model according to any one of claims 1-6.

9. A computer-readable storage medium having stored thereon a computer program, characterized in that The computer program is executed by the processor to implement the method for predicting the running state of a physical entity based on a digital twin model according to any one of claims 1-6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the method for predicting the running state of a physical entity based on a digital twin model according to any one of claims 1-6.

Images