Device model upgrade method, device, device, server and storage medium

Abstract

The present invention discloses a device model upgrade method, apparatus, equipment, server and storage medium, wherein the device model upgrade method comprises: each device obtains the training data of its own model to be updated; the model to be updated is trained using the training data of each device to obtain the optimal model after training; and the optimal model is upgraded to each device. In this way, the autonomous upgrade of the model is realized, which does not require human participation and can reduce the labor cost during the upgrade.

Description

Technical Field

[0001] The present invention relates to the field of cloud computing applications, and in particular to a device model upgrading method, apparatus, equipment, server and storage medium. Background Art

[0002] Currently, research on deep learning and machine learning is booming and producing results frequently, which requires higher and higher technical levels and lower human participation.

[0003] In the software industry, rapid iterative development models such as agile development mean that version upgrades are becoming more and more frequent. With the increase in computing power and distributed cluster deployment, systems are becoming larger and larger. Existing systems require manpower to trigger upgrades and wait for upgrade results when upgrading models, which has low operability. It is becoming increasingly important to train, release, and upgrade models quickly, autonomously, and without affecting business. Summary of the invention

[0004] The present invention provides a device model upgrade method, apparatus, equipment, server and storage medium, which can realize autonomous upgrade of the model without manual triggering.

[0005] To solve the above technical problems, the first technical solution provided by the present invention is: to provide a model upgrade method for a device, comprising: each device obtains the training data of its own model to be updated; the model to be updated is trained using the training data of each of the devices to obtain the optimal model after training; and the optimal model is upgraded to each of the devices.

[0006] Among them, the device includes a first device and a second device; the step of using the training data of each device to train the model to be updated to obtain the optimal model after training includes: the first device and the second device respectively train the model to be updated through their respective training data to obtain the trained first model; the first device receives the first model trained by the second device; the first model trained by the first device and the first model trained by the second device are compared to select the optimal model.

[0007] Among them, the step of using the training data of each of the devices to train the model to be updated to obtain the optimal model after training includes: summarizing the training data of each of the devices; training the model to be updated with the summarized training data to obtain the optimal model.

[0008] Among them, the step of aggregating the training data of each of the devices includes: each of the devices sends its own training data to the central server respectively; the step of training the model to be updated by using the aggregated training data to obtain the optimal model includes: training the model to be updated by using the training data of each of the devices through the central server to obtain the optimal model.

[0009] Wherein, it is characterized in that the step of upgrading the optimal model to each of the devices includes: sending the optimal model to each of the devices / the second device through the central server and / or the first device.

[0010] Among them, sending the optimal model to each of the devices / the second devices through the central server and / or the first device includes: sending the optimal model to each of the devices / the second devices by adopting any one or any combination of grayscale release, rolling release, blue-green release.

[0011] Among them, sending the optimal model to each of the devices / the second devices in a grayscale release manner includes: selecting some devices from the devices / the second devices as pilot devices, and sending the optimal model to the pilot devices for testing; after the test results are stable, sending the optimal model to the remaining devices except the pilot devices.

[0012] Among them, the upgrading of the optimal model to each of the devices includes: starting to update the optimal model in the device, and temporarily using the model to be updated for data testing during the process of starting to update the optimal model; after the optimal model is started and updated, the historical data in the model to be updated is sent to the optimal model, and the model to be updated is taken offline, and the optimal model is used for data testing; or the historical data in the model to be updated of the device is sent to the conversion device, and the model to be updated is taken offline; starting to update the optimal model in the device, and after the optimal model is started and updated, the historical data returned by the conversion device is received, and the optimal model is used for data testing.

[0013] To solve the above technical problems, the second technical solution provided by the present invention is: to provide a device, comprising: a training data acquisition module, used to obtain the training data of each model to be updated; a training module, used to train the model to be updated using the training data of each of the devices to obtain the optimal model after training; an update module, used to upgrade the optimal model to each of the devices.

[0014] To solve the above technical problems, the third technical solution provided by the present invention is: to provide a model upgrade method for a device, comprising: a central server receives training data sent by each of the devices; the central server uses the training data to train the loaded model to be updated to obtain the optimal model after training; the central server upgrades the optimal model to each of the devices.

[0015] To solve the above technical problems, the fourth technical solution provided by the present invention is: to provide a server, comprising: a training data acquisition module, used to receive the training data sent by each of the devices; a training module, used to use the training data to train the loaded model to be updated to obtain the optimal model after training; an update module, used to upgrade the optimal model to each of the devices.

[0016] To solve the above technical problems, the fifth technical solution provided by the present invention is: to provide a model upgrade device for a device, comprising: a memory and a processor, wherein the memory stores program instructions, and the processor calls the program instructions from the memory to execute the model upgrade method for the device described in any one of the above items.

[0017] In order to solve the above technical problems, the fifth technical solution provided by the present invention is: providing a computer-readable storage medium storing a program file, wherein the program file can be executed to implement the model upgrade method of the device described in any one of the above items.

[0018] The beneficial effect of the present invention is different from the prior art. The present invention enables each device to obtain the training data of its own model to be updated; the model to be updated is trained using the training data of each device to obtain the optimal model after training; and the optimal model is upgraded to each device. In this way, the autonomous upgrade of the model is realized, which does not require human participation and can reduce the labor cost during the upgrade. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the 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 work, among which:

[0020] Figure 1 is a flow chart of a first embodiment of a model upgrading method for a device of the present invention;

[0021] Figure 2a for Figure 1 A structural diagram of an embodiment of step S12;

[0022] Figure 2b for Figure 1 A schematic flow chart of an embodiment of step S12;

[0023] Figure 3a for Figure 1 A schematic structural diagram of another embodiment of step S12;

[0024] Figure 3b for Figure 1 A schematic flow chart of another embodiment of step S12;

[0025] Figure 4a for Figure 1 A schematic flow chart of a first embodiment of step S13;

[0026] Figure 4b for Figure 1 A schematic structural diagram of a first embodiment of step S13;

[0027] Figure 5 for Figure 1 A schematic structural diagram of a second embodiment of step S13;

[0028] Figure 6 for Figure 1 A schematic structural diagram of a third embodiment of step S13;

[0029] Figure 7 It is a structural schematic diagram of an embodiment of the device of the present invention;

[0030] Figure 8 A schematic flow chart of a second embodiment of a model upgrading method for a device of the present invention;

[0031] Fig. 9 A schematic diagram of the structure of an embodiment of the central server of the present invention;

[0032] Fig.10 A schematic structural diagram of an embodiment of a model upgrading device of the present invention;

[0033] Fig.11 It is a schematic diagram of the structure of the computer-readable storage medium of the present invention. DETAILED DESCRIPTION

[0034] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0035] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0036] See also Figure 1 , is a flow chart of the first embodiment of the model upgrading method of the device of the present invention. It includes:

[0037] Step S11: Each device obtains the training data of its own model to be updated.

[0038] Specifically, when performing a model upgrade, the device collects training data. In one embodiment, the training data includes historical sample data of the model to be updated and current test data. Specifically, the historical sample data is the data involved in the historical test process of the model to be updated, and the current test data is the data currently obtained by the model to be updated but not yet tested.

[0039] Step S12: Use the training data of each device to train the model to be updated to obtain the optimal model after training.

[0040] Specifically, in this embodiment, when the model is upgraded, the device can upgrade autonomously, and can also be upgraded uniformly through a central server.

[0041] Please refer to Figure 2a and Figure 2b In this embodiment, the device autonomously trains and upgrades the model, specifically including:

[0042] Step S121: the first device and the second device respectively train the model to be updated using their respective training data to obtain a trained first model.

[0043] like Figure 2a As shown, in this embodiment, for example, multiple devices include device 1, device 2, device 3 and device 4, and devices 1, device 2, device 3 and device 4 have communication transmission functions. In one embodiment, device 1, device 2, device 3 and device 4 are the same device, for example, they can be cameras, speakers, etc. at the same time, and there is no specific limitation. Or, in another embodiment, if device 1, device 2, device 3 and device 4 are cameras, such as monitoring devices, they must be monitoring devices in the same area, and they are loaded with the same model to be updated.

[0044] In this embodiment, device 1, device 2, device 3 and device 4 are divided into a first device 21 and a second device 22. Specifically, Figure 2aAs shown, device 1 is the first device 21, and devices 2, 3, and 4 are the second devices 22. When upgrading the model, the first device 21 (device 1) and the second device 22 (device 2, device 3, and device 4) respectively obtain their own corresponding training data, and train the models to be updated loaded thereon through their respective training data to obtain the trained first model.

[0045] Step S122: The first device receives the first model trained by the second device.

[0046] Specifically, after the first device 21 and the second device 22 use the training data to train the model to be updated to obtain the first model, the second device 22 (device 2, device 3 and device 4) sends its own first model to the first device 21 (device 1), and the first device 21 receives the first model trained by the second device to test the trained first model.

[0047] Step S123: Compare the first model trained by the first device and the first model trained by the second device to select the optimal model.

[0048] Specifically, after receiving the first model trained by the second device, the first device 21 tests the trained first model. Specifically, during the test, the first device 21 obtains a labeled sample data, tests the sample data in the first model corresponding to each device, obtains the test result, compares the test result of each first model with the labeled result, and selects the best model from each first model.

[0049] In another embodiment, the model can also be upgraded through the central server. Figure 3a and Figure 3b ,include:

[0050] Step S124: Summarize the training data of each device.

[0051] Specifically, Figure 3a As shown, for example, device 31 includes devices 1 to 6, and devices 1 to 6 have a communication transmission function with a central server 32. Each device 31 sends its own training data to the central server 32 to aggregate the training data to the central server. It should be noted that the central server 32 is loaded with the same model to be updated as devices 1 to 6.

[0052] Step S125: Train the model to be updated using the summarized training data to obtain the optimal model.

[0053] After receiving the aggregated training data, the central server 32 uses the training data to train the model to be updated that is the same as that of the device 31 and is loaded in the central server 32. The model obtained after training is the optimal model.

[0054] Step S13: Upgrade the optimal model to each device.

[0055] Specifically, Figure 2a and Figure 2b In the described embodiment, after the first device 21 (device 1) selects the optimal model, the optimal model is upgraded to each device.

[0056] For example, in one embodiment, if the optimal model selected by the first device 21 (device 1) is the first model trained by device 1, device 1 sends its own first model to the second device 22 (device 2, device 3 and device 4); or in another embodiment, if the optimal model selected by the first device 21 (device 1) is the first model trained by device 2, device 1 sends the first model corresponding to device 2 to device 1, device 3 and device 4; or in another embodiment, if the optimal model selected by the first device 21 (device 1) is the first model trained by device 2, device 1 feeds back the comparison result to device 2, and device 2 sends the corresponding first model to device 3 and device 4. Since device 1 has obtained the first model of device 2, there is no need to resend it.

[0057] exist Figure 3a and Figure 3b In the illustrated embodiment, after the central server 32 obtains the optimal model after training, it sends the optimal model to each device 31.

[0058] In one embodiment, the optimal model can be upgraded to each device by using any one or any combination of grayscale release, rolling release, blue-green release to send the optimal model to each device / the second device. Figure 2a and Figure 2b In the embodiment shown, device 1 can send the optimal model to each second device by using any one or any combination of grayscale release, rolling release, blue-green release, etc. Figure 3a and Figure 3b In the illustrated embodiment, the central server 32 may send the optimal model to each device 31 by using any one of grayscale release, rolling release, blue-green release, or any combination thereof.

[0059] Specifically, in one embodiment, grayscale release is used as an example for explanation. Figure 4a ,include:

[0060] Step S41: Select some devices from the device / second device as pilot devices, and send the optimal model to the pilot devices for testing.

[0061] Specifically, Figure 4b As shown, the devices include pilot devices 41 (pilot device 1, pilot device 2, pilot device 3) and other devices 42 (device 4, device 5). The central server 43 sends the optimal model to the pilot devices 41 (pilot device 1, pilot device 2, pilot device 3) first, and performs a pre-test in the pilot devices 41, while the other devices 42 (device 4, device 5) still use the model to be updated.

[0062] Step S42: After the test results are stable, the optimal model is sent to the remaining devices except the pilot device.

[0063] After the pilot devices 41 (pilot device 1 , pilot device 2 , pilot device 3 ) are tested using the optimal model and the test results are stable, the central server 43 sends the optimal model to the remaining devices 42 except the pilot device 41 .

[0064] Figure 4a and Figure 4b The embodiment of model upgrade by the central server is described. In another embodiment, the device autonomously performs model upgrade. When the model is released, the same method as above can be adopted to first select the pilot device for update and then update other devices. The details will not be repeated here.

[0065] In another embodiment, rolling release can also be adopted, and the rolling release method is to release one by one, that is, the central server first sends the optimal model to device 1, tests it in device 1, and then sends the optimal model to device 2 after the test is completed, until it is sent to the last device. The blue-green release method is to divide all devices into two types, for example, they can be divided according to device type and device location, and the model of one type of device is upgraded first, and then the model of the other type of device is upgraded.

[0066] In one embodiment, when updating the model in the device, it is necessary to consider whether the device resources are sufficient. When the device resources are sufficient, that is, the storage space is large enough, the Figure 5 The model is upgraded in the following way, including:

[0067] Step S51: starting to update the optimal model in the device, and in the process of starting to update the optimal model, temporarily using the model to be updated to perform data testing.

[0068] Taking the central server update as an example, specifically, the central server sends the optimal model to the device, starts the update in the device, and during the update process, the model to be updated remains unchanged, and the model to be updated is temporarily used for data testing.

[0069] Taking the device's autonomous update as an example, after the device obtains the optimal model, the update is started in the device. During the update process, the model to be updated remains unchanged, and the model to be updated is temporarily used for data testing.

[0070] Step S52: After the optimal model is started and updated, the historical data in the model to be updated is sent to the optimal model, and the model to be updated is taken offline, and the optimal model is used for data testing.

[0071] After the optimal model is started and updated, the historical data in the model to be updated is sent to the optimal model. At this time, the model to be updated is taken offline, and then the optimal model can be used for data testing.

[0072] In another embodiment, if the device resources are insufficient, that is, the storage space is not large enough, the Figure 6 The model is upgraded in the following way, including:

[0073] Step S61: Send the historical data in the device's model to be updated to other devices, and take the model to be updated offline.

[0074] Specifically, a conversion device may be introduced to send historical data in the model to be updated of the device to the conversion device and pre-store the data in the conversion device. After the data is sent, the model to be updated is taken offline.

[0075] Step S62: Start updating the optimal model in the device, and after the optimal model is started and updated, receive the historical data returned by other devices, and use the optimal model to perform data testing.

[0076] After the central server or device sends the optimal model to the device, the device starts updating the optimal model. After the optimal model is started and updated, the historical data returned from the conversion device is received, and then the optimal model can be used for data testing.

[0077] The model upgrade method for the device provided in this embodiment can be updated autonomously through the device or the central server without manual triggering, which can reduce the manpower cost during the upgrade and is sufficient; and when released, it can realize gradual upgrade switching according to existing resources without affecting existing business.

[0078] See also Figure 7 , is a schematic diagram of the structure of an embodiment of the device of the present invention, which includes: a training data acquisition module 71, a training module 72 and an updating module 73.

[0079] The training data acquisition module 71 is used to acquire the training data of each model to be updated.

[0080] The training module 72 is used to train the model to be updated using the training data of each device to obtain the optimal model after training.

[0081] The updating module 73 is used to update the optimal model to each of the devices.

[0082] See also Figure 8 , is a flow chart of a second embodiment of a model upgrading method for a device of the present invention, which comprises:

[0083] Step S81: The central server receives the training data sent by each of the devices.

[0084] Specifically, each device sends the training data to the central server, and the central server receives the training data sent by each device.

[0085] Step S82: The central server uses the training data to train the loaded model to be updated to obtain the optimal model after training.

[0086] After receiving the aggregated training data, the central server 32 uses the training data to train the model to be updated that is the same as that of the device 31 and is loaded in the central server 32. The model obtained after training is the optimal model.

[0087] Step S83: The central server upgrades the optimal model to each of the devices.

[0088] After the central server obtains the optimal model after training, it sends the optimal model to each device. In one embodiment, the optimal model can be upgraded to each device by using any one or any combination of grayscale release, rolling release, blue-green release to send the optimal model to each of the devices / the second device.

[0089] The model upgrade method for the device provided in this embodiment can be updated autonomously through the device or the central server without manual triggering, which can reduce the manpower cost during the upgrade and is sufficient; and when released, it can realize gradual upgrade switching according to existing resources without affecting existing business.

[0090] See also Fig. 9 , is a schematic diagram of the structure of an embodiment of the server of the present invention, which includes: a training data acquisition module 91, a training module 92 and an updating module 93.

[0091] The training data acquisition module 91 is used to receive the training data sent by each of the devices.

[0092] The training module 92 is used to train the loaded model to be updated using the training data to obtain the optimal model after training.

[0093] The updating module 93 is used to update the optimal model to each of the devices.

[0094] See also Fig.10 , is a schematic diagram of a structure of an embodiment of a model upgrading device of a device of the present invention. The model upgrading device of the device comprises a memory 102 and a processor 101 which are connected to each other.

[0095] The memory 102 is used to store program instructions for implementing any of the above-mentioned device model upgrade methods.

[0096] The processor 101 is used to execute program instructions stored in the memory 102 .

[0097] The processor 101 may also be referred to as a CPU (Central Processing Unit). The processor 101 may be an integrated circuit chip having signal processing capabilities. The processor 101 may also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0098] The memory 102 can be a memory stick, a TF card, etc., which can store all the information in the model upgrade device of the device, including the input raw data, computer programs, intermediate operation results and final operation results are all stored in the memory. It stores and retrieves information according to the location specified by the controller. With the memory, the model upgrade device of the device has a memory function and can ensure normal operation. The memory of the model upgrade device of the device can be divided into main memory (internal memory) and auxiliary memory (external memory) according to the purpose of the memory, and there is also a classification method of dividing it into external memory and internal memory. External memory is usually a magnetic medium or an optical disk, etc., which can store information for a long time. Memory refers to the storage component on the motherboard, which is used to store the data and programs currently being executed, but is only used to temporarily store programs and data. If the power is turned off or the power is cut off, the data will be lost.

[0099] In the several embodiments provided in the present application, it should be understood that the disclosed methods and devices can be implemented in other ways. For example, the device implementation described above is only schematic. For example, the division of modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0100] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0101] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0102] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a system server, or a network device, etc.) or a processor to execute all or part of the steps of the various implementation methods of the present application.

[0103] See also Fig.11, which is a schematic diagram of the structure of the computer-readable storage medium of the present invention. The storage medium of the present application stores a program file 201 that can implement the model upgrade device method of all the above-mentioned devices, wherein the program file 201 can be stored in the above-mentioned storage medium in the form of a software product, including a number of instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) or a processor (processor) to execute all or part of the steps of each implementation method of the present application. The aforementioned storage device includes: various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, or a terminal device such as a computer, a server, a mobile phone, and a tablet.

[0104] The above are only implementation modes of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A device model upgrade method, characterized in that: include: Each device obtains the training data of its own model to be updated; Using the training data of each of the devices to train the model to be updated, to obtain the optimal model after training; Upgrading the optimal model to each of the devices; The device comprises a first device and a second device; The step of training the model to be updated by using the training data of each device to obtain the optimal model after training comprises: The first device and the second device respectively train the model to be updated using their respective training data to obtain a trained first model; The first device receives the first model trained by the second device; Compare the first model trained by the first device and the first model trained by the second device, and select the optimal model; The training data includes the historical sample data of the model to be updated and current test data. The historical sample data is the data involved in the historical test process of the model to be updated, and the current test data is the data currently obtained by the model to be updated but not yet tested.

2. The device model upgrade method according to claim 1, characterized in that: The step of training the model to be updated by using the training data of each device to obtain the optimal model after training comprises: Aggregating the training data of each of the devices; The model to be updated is trained using the summarized training data to obtain the optimal model.

3. The model upgrading method of the device according to claim 2, characterized in that: The step of aggregating the training data of each of the devices comprises: Each of the devices sends its own training data to the central server respectively; The step of training the model to be updated by using the summarized training data to obtain the optimal model comprises: The central server uses the training data of each of the devices to train the model to be updated to obtain the optimal model.

4. The device model upgrade method according to claim 3, characterized in that: The step of upgrading the optimal model to each of the devices comprises: Sending the optimal model to each of the devices through the central server; or The optimal model is sent to each of the second devices through the first device.

5. The model upgrading method of the device according to claim 4, characterized in that: The sending the optimal model to each of the devices through the central server comprises: Sending the optimal model to each of the devices by adopting any one of grayscale release, rolling release, blue-green release, or any combination thereof; or The sending the optimal model to each of the second devices through the first device comprises: The optimal model is sent to each of the second devices by adopting any one of grayscale release, rolling release, blue-green release or any combination thereof.

6. The device model upgrade method according to claim 5, characterized in that: The method of sending the optimal model to each of the devices / the second device in a grayscale publishing manner includes: Selecting some devices from the device / the second device as pilot devices, and sending the optimal model to the pilot devices for testing; After the test results are stable, the optimal model is sent to the remaining devices except the pilot device.

7. The device model upgrade method according to claim 1, characterized in that: The step of upgrading the optimal model to each of the devices comprises: Initiating updating of the optimal model in the device, and in the process of initiating updating of the optimal model, temporarily using the model to be updated to perform data testing; After the optimal model is started and updated, the historical data in the model to be updated is sent to the optimal model, and the model to be updated is offline, and the optimal model is used for data testing; or Sending historical data in the model to be updated of the device to a conversion device, and taking the model to be updated offline; The optimal model is started to be updated in the device, and after the optimal model is started and updated, the historical data returned by the conversion device is received, and the optimal model is used to perform data testing.

8. A computer device, characterized in that: include: A training data acquisition module is used to obtain the training data of each model to be updated; A training module, used to train the model to be updated using the training data of each device to obtain the optimal model after training; An updating module, used for updating the optimal model to each of the devices; The device comprises a first device and a second device; The training module trains the model to be updated using the training data of each device, and the step of obtaining the trained optimal model includes: The first device and the second device respectively train the model to be updated using their respective training data to obtain a trained first model; The first device receives the first model trained by the second device; Compare the first model trained by the first device and the first model trained by the second device, and select the optimal model; The training data includes the historical sample data of the model to be updated and current test data. The historical sample data is the data involved in the historical test process of the model to be updated, and the current test data is the data currently obtained by the model to be updated but not yet tested.

9. A device for upgrading a model of an equipment, characterized in that: include: A memory and a processor, wherein the memory stores program instructions, and the processor retrieves the program instructions from the memory to execute the model upgrading method of the device according to any one of claims 1-7.

10. A computer-readable storage medium, characterized in that: A program file is stored, and the program file can be executed to implement the model upgrading method of the device according to any one of claims 1 to 7.

Images