Semi-empirical semi-physical vehicle longitudinal model building method, terminal and medium

By constructing a semi-experience and semi-physical vehicle longitudinal model and using neural network model to process high-frequency modules, the step length problem of the vehicle longitudinal physical model in the HIL system is solved, and the simulation efficiency and reliability are improved.

CN118940414APending Publication Date: 2024-11-12SINO TRUK JINAN POWER CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411136074.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-19
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

When building a longitudinal physical model of the vehicle in the prior art, the model solution step length is much smaller than the fixed step length, which makes it impossible to solve in the HIL real-time system, and the lack of physical theory support for simplifying the model or changing parameters, affecting the reliability of the simulation results.

Method used

By selecting a sub-physical model with a higher number of high-frequency time intervals than the threshold, combining the neural network model, removing the high-frequency module, and building a semi-empirical semi-physical vehicle longitudinal model to realize fixed step size solution while retaining some physical characteristics.

Benefits of technology

It realizes fixed step size solution in real-time system, ensures simulation efficiency, accuracy and reliability of results, and meets the requirements of HIL testing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118940414A_ABST
    Figure CN118940414A_ABST
Patent Text Reader

Abstract

The invention relates to the field of whole vehicle model simulation, and particularly discloses a semi-empirical semi-physical whole vehicle longitudinal model building method, a terminal and a medium, and the method comprises the following steps: building a whole vehicle longitudinal physical model; simulating the whole vehicle longitudinal physical model, screening out sub-physical models of which the number of high-frequency intervals is greater than a threshold value, and recording the sub-physical models as to-be-processed sub-physical models; constructing a neural network model, and selecting a target parameter of the to-be-processed sub-physical model as an input / output quantity of the neural network model; selecting target parameters under a plurality of working conditions simulated by the whole vehicle longitudinal physical model to construct a sample set to train the neural network model, wherein the trained neural network model is an empirical model of the to-be-processed sub-physical model; and importing the obtained empirical model into an original vehicle longitudinal physical model to replace the original to-be-processed sub-physical model, and generating a semi-empirical semi-physical vehicle longitudinal model. According to the method, a high-frequency module is removed, the solving condition of a real-time system is met, and the simulation efficiency, precision and result reliability are guaranteed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of whole vehicle model simulation, and in particular to a semi-empirical and semi-physical whole vehicle longitudinal model building method, terminal and medium. Background Art

[0002] With the continuous development of the automotive industry, the functions of on-board controllers are gradually increasing, and the testing requirements are also gradually increasing. HIL (Hardware in the Loop) testing can reduce a lot of manpower and expenses for testing, and is essential for HIL verification of the longitudinal dynamics model of the vehicle for the power domain controller. As a physical model modeling software, Amesim can well reflect the dynamic characteristics of the model, but it will lead to many cases in the solution process where the model solution step is much smaller than the fixed step, especially for models with more complex modules such as the longitudinal physical model of the vehicle. The model is difficult to solve during fixed step simulation, resulting in the inability to solve the model when deployed in real-time systems such as HIL real-time machines.

[0003] Usually, the original longitudinal physical model of the whole vehicle is simplified or the parameters are changed so that the model can be solved with a fixed step size. The disadvantage is that the changeable part of the simplified model is small and has little impact on the solution step size. Most of the changed parameters are based on random attempts. There is no physical theory to support the new parameters, which affects the reliability of the simulation results. Summary of the invention

[0004] To solve the above problems, the present invention provides a method, terminal and medium for building a semi-empirical and semi-physical whole vehicle longitudinal model, which selects a sub-physical model whose number of high-frequency time intervals is greater than a threshold to build an empirical model, and then builds a semi-empirical and semi-physical whole vehicle longitudinal model. The high-frequency module is removed, and a fixed-step solution can be performed to meet the solution conditions of the real-time system. At the same time, some physical characteristics are retained, so that the model basically conforms to the physical logic, ensuring the simulation efficiency, accuracy and reliability of the results.

[0005] In a first aspect, the technical solution of the present invention provides a semi-empirical and semi-physical method for building a longitudinal model of a whole vehicle, which is implemented based on the Amesim physical model building software and includes the following steps: Building a longitudinal physical model of the whole vehicle, which includes multiple sub-physical models; Simulate the longitudinal physical model of the whole vehicle, and select the sub-physical models whose number of high-frequency intervals is greater than the threshold value, and record them as the sub-physical models to be processed; Constructing a neural network model, and selecting target parameters of the sub-physical model to be processed as input and output quantities of the neural network model; Select target parameters under several working conditions of the longitudinal physical model simulation of the whole vehicle to construct a sample set to train the neural network model, and the trained neural network model is the empirical model of the sub-physical model to be processed; The obtained empirical model is imported into the original longitudinal physical model of the whole vehicle to replace the original sub-physical model to be processed, so as to generate a semi-empirical and semi-physical longitudinal model of the whole vehicle.

[0006] In an optional embodiment, when simulating the longitudinal physical model of the entire vehicle, the simulation type is parallel operation, the simulation solution method is variable step size solution, and four simulation conditions are used, one of which is a condition made from data collected from the actual vehicle, and the other three are standard conditions.

[0007] In an optional implementation, screening out the sub-physical models whose number of high-frequency intervals is greater than a threshold value specifically includes: Select the frequency option in the performance analysis options, click the run button to display the high-frequency time interface, and select the sub-physical model whose number of high-frequency intervals is greater than the threshold through the high-frequency interval information on the high-frequency time interface.

[0008] In an optional implementation, the screened sub-physical model to be processed is an engine module.

[0009] In an optional embodiment, the target parameters of the sub-physical model to be processed include the engine acceleration signal and the speed of the engine fed back to the engine by the engine output shaft as inputs to the neural network model, and the engine output torque, engine speed, engine brake mean effective pressure, engine maximum torque and engine minimum torque as outputs of the neural network model.

[0010] In an optional embodiment, the constructed neural network model includes two hidden layers, wherein the first layer is a recurrent neural network, the second layer is a linear layer, the number of cell units in the two layers is 9 to 11, and the number of model iterations is less than 400.

[0011] In an optional implementation, the obtained empirical model is imported into the original longitudinal physical model of the whole vehicle to replace the original sub-physical model to be processed, and a semi-empirical and semi-physical longitudinal model of the whole vehicle is generated, which specifically includes: After the neural network model training is completed, the FMU file is generated in the ROM creation program; Delete the sub-physical model to be processed in the original longitudinal physical model of the whole vehicle; In the FMU import assistant option, select the FMU file generated in the ROM creation program. In the module library, drag the neural network model into the view, connect the input and output to complete the replacement of the sub-physical model to be processed, and generate a semi-empirical and semi-physical vehicle longitudinal model.

[0012] In an optional embodiment, after the semi-empirical and semi-physical vehicle longitudinal model is generated, the following steps are further included: The semi-empirical and semi-physical vehicle longitudinal model is simulated, and the simulation solution type is fixed step solution, and the simulation step is 0.001s; In the Create Interface Block option, select the module type as FMI protocol and connect the output of the vehicle longitudinal physical model to the Create Interface Block option; Click the FMU Import Assistant option and select the export type to generate an FMU file of the semi-empirical and semi-physical vehicle longitudinal model.

[0013] In a second aspect, the technical solution of the present invention provides a terminal, including: A memory for storing a semi-empirical and semi-physical vehicle longitudinal model building program; A processor is used to implement the steps of any of the above-mentioned semi-empirical and semi-physical vehicle longitudinal model building methods when executing the semi-empirical and semi-physical vehicle longitudinal model building program.

[0014] In a third aspect, the technical solution of the present invention provides a computer-readable storage medium, on which is stored a semi-empirical and semi-physical whole vehicle longitudinal model building program, which, when executed by a processor, implements the steps of the semi-empirical and semi-physical whole vehicle longitudinal model building method as described in any of the above items.

[0015] The semi-empirical and semi-physical vehicle longitudinal model building method, terminal and medium provided by the present invention have the following beneficial effects compared with the prior art: a sub-physical model with a number of high-frequency time intervals greater than a threshold is selected to construct an empirical model, and then a semi-empirical and semi-physical vehicle longitudinal model is constructed. Compared with the original model, the high-frequency module is removed, and a fixed-step solution can be performed to meet the solution conditions of the real-time system. At the same time, some physical properties are retained, so that the model basically conforms to physical logic, which is convenient for researchers to better understand the vehicle longitudinal physical model. The trained modules are all smaller sub-modules, and their inputs and outputs are mostly strongly correlated at the physical level, which can ensure the accuracy of the model trained by the neural network, and ensure the simulation efficiency, accuracy and reliability of the results. BRIEF DESCRIPTION OF THE DRAWINGS In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 It is a schematic flow chart of a semi-empirical and semi-physical vehicle longitudinal model building method provided by an embodiment of the present invention.

[0017] Figure 2 It is a schematic diagram of the structure of a terminal provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0018] In order to enable those skilled in the art to better understand the scheme of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0019] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art of the present invention. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0020] Figure 1 This is a flow chart of a semi-empirical and semi-physical vehicle longitudinal model building method provided by an embodiment of the present invention. The method is implemented based on Amesim physical model building software and executed by a computer device. According to different requirements, the order of the steps in the flow chart can be changed, and some can be omitted.

[0021] like Figure 1 As shown, the method includes the following steps.

[0022] S1, building a longitudinal physical model of the whole vehicle, which includes multiple sub-physical models.

[0023] First, you need to build a longitudinal physical model of the whole vehicle in Amesim, which includes the engine model, gearbox model, ECU (Engine Control Unit), TCU (Transmission Control Unit), and vehicle model. This model is a physical model that can reflect the real physical properties.

[0024] S2, simulate the longitudinal physical model of the whole vehicle, screen out the sub-physical models whose number of high-frequency intervals is greater than the threshold, and record them as the sub-physical models to be processed.

[0025] The original vehicle longitudinal physical model constructed in step S1 is simulated. At this time, the simulation type is parallel operation, and the simulation solution method is variable step size solution. Four simulation conditions are used in the simulation. The first one is the condition made of data collected from the actual vehicle, and the last three are standard conditions, namely FTP72 (Federal Test Procedure-72), HWFET (Highway FuelEconomy Test), and NEDC (New European Driving Cycle).

[0026] After the simulation is completed, select the sub-physical models whose number of high-frequency intervals is greater than the threshold, including selecting the Frequencies option in the Performance Analyzer option, clicking the Run button to display the high-frequency time interface, and selecting the sub-physical models whose number of high-frequency intervals is greater than the threshold through the high-frequency interval information on the high-frequency time interface.

[0027] Specifically, click Performance Analyzer on the simulation page, select Frequencies, and then click Start to see the high-frequency time appearing in the model in the right view. This part can analyze the dynamic characteristics of the longitudinal physical model of the entire vehicle and find out which modules contain high-frequency intervals. The solution step size is much smaller than the fixed step size.

[0028] S3, constructing a neural network model, and selecting target parameters of the sub-physical model to be processed as input and output quantities of the neural network model.

[0029] In this embodiment, the high-frequency module is converted into an empirical model to replace the original physical model, and the low-frequency module does not need to be converted, thereby constructing a semi-empirical and semi-physical model. The empirical model is constructed using a neural network.

[0030] After screening in step S2, it is found that the high-frequency modules are concentrated in the engine module. The engine module is taken as the sub-physical model to be processed, and the target parameters of the engine module are taken as the input and output of the neural network model.

[0031] In the parameter interface of the software, select the quantities that need to be observed as the input and output of the neural network. When selecting the input and output, select fewer but more important input and output quantities for training based on the actual vehicle data and previous simulation experience to ensure the accuracy of the empirical model. In this embodiment, the acceleration signal (load control) sent by the ECU to the engine and the speed (velcoity) fed back to the engine by the engine output shaft are selected as the input quantities of the neural network, and the engine output torque (torque), engine speed (engine speed), engine brake mean effective pressure (engine brake meaneffective pressure), engine maximum torque (maximum engine torque) and engine minimum torque (minimum engine torque) are selected as the output quantities of the neural network.

[0032] The neural network model constructed in this embodiment includes two hidden layers, wherein the first layer is a recurrent neural network, and the second layer is a linear layer. The number of cell units in the two layers is 9 to 11, and the number of model iterations is less than 400.

[0033] S4, selecting target parameters under several working conditions of the longitudinal physical model simulation of the whole vehicle to construct a sample set to train the neural network model, and the trained neural network model is the empirical model of the sub-physical model to be processed.

[0034] After selecting the input and output, the target parameters under several working conditions of the longitudinal physical model simulation of the whole vehicle are selected to construct a sample set to train the neural network model. Specifically, the ROM Builder app tool provided by Amesim is used to select the above four working conditions to form the training set.

[0035] After the training is completed, the trained neural network model is saved. The newly generated neural network model no longer has the original physical properties. The new model reflects the mapping relationship of the data and is an empirical model.

[0036] S5, importing the obtained empirical model into the original longitudinal physical model of the whole vehicle to replace the original sub-physical model to be processed, and generating a semi-empirical and semi-physical longitudinal model of the whole vehicle.

[0037] To complete the model replacement, after the neural network model training is completed, an FMU file is generated in the ROM creation program; then the sub-physical model to be processed in the original vehicle longitudinal physical model is deleted; in the FMU import assistant option, the FMU file generated in the ROM creation program is selected, and the neural network model appears in the module library. Drag-in view, connect the input and output to complete the replacement of the sub-physical model to be processed, and generate a semi-empirical and semi-physical vehicle longitudinal model.

[0038] Specifically, directly generate the FMU file in the ROM Builder app, re-import the generated FMU file into the original model to replace the original physical module. The operation is as follows: first delete the original engine model and then select FMU importassistant, select the FMU file just generated, and a new neural network model will appear in the module library on the right. Drag it into the view and connect the input and output to complete the replacement.

[0039] In order to realize the HIL test based on the semi-empirical and semi-physical full vehicle longitudinal model, the following steps are performed after the semi-empirical and semi-physical full vehicle longitudinal model is generated.

[0040] S6, simulate the semi-empirical and semi-physical vehicle longitudinal model, the simulation solution type is fixed step solution, and the simulation step is 0.001s.

[0041] S7, select the module type as FMI protocol in the Create Interface Block option, and connect the output of the vehicle longitudinal physical model to the Create Interface Block option.

[0042] Specifically, click create interface block interface model in the software. This model is used for joint simulation of Amesime model and other simulation software. Select the module type as Functional Mock-up Interface (FMI) protocol. The input of the overall model is the start signal output by the driver to ECU, the acceleration signal to ECU and TCU, the brake signal to the wheel end and the gear shift signal to TCU. The output of the overall model is the current vehicle speed, the current gear position of the gearbox and the engine speed. Connect these original outputs to the block interface module.

[0043] S8, click the FMU Import Assistant option, select the export type to generate the FMU file of the semi-empirical and semi-physical vehicle longitudinal model.

[0044] Click the FMU import assistant option in the software and select the export type as Co-simuliation2.0 to generate an overall semi-empirical and semi-physical FMU file. When HIL testing is required, import the FMU file into the HIL operating software (such as CongurfationDesk software and EM software) to perform HIL testing.

[0045] The reconstructed model in this embodiment is a semi-empirical and semi-physical model. Compared with the original model, the high-frequency module is removed and a fixed-step solution can be performed to meet the solution conditions of the real-time system. At the same time, some physical properties are retained, so that the model basically conforms to physical logic, which is convenient for researchers to better understand the longitudinal physical model of the whole vehicle.

[0046] Figure 2 A schematic diagram of the structure of a terminal 200 provided in an embodiment of the present invention includes: a processor 210, a memory 220 and a communication unit 230. The processor 210 is used to implement the following steps when implementing the semi-empirical and semi-physical vehicle longitudinal model building program stored in the memory 220: Based on Amesim physical model modeling software, it includes the following steps: Building a longitudinal physical model of the whole vehicle, which includes multiple sub-physical models; Simulate the longitudinal physical model of the whole vehicle, and select the sub-physical models whose number of high-frequency intervals is greater than the threshold value, and record them as the sub-physical models to be processed; Constructing a neural network model, and selecting target parameters of the sub-physical model to be processed as input and output quantities of the neural network model; Select target parameters under several working conditions of the longitudinal physical model simulation of the whole vehicle to construct a sample set to train the neural network model, and the trained neural network model is the empirical model of the sub-physical model to be processed; The obtained empirical model is imported into the original longitudinal physical model of the whole vehicle to replace the original sub-physical model to be processed, so as to generate a semi-empirical and semi-physical longitudinal model of the whole vehicle.

[0047] The terminal 200 includes a processor 210, a memory 220 and a communication unit 230. These components communicate via one or more buses. It will be appreciated by those skilled in the art that the server structure shown in the figure does not limit the present invention, and it may be a bus structure or a star structure, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0048] The memory 220 can be used to store the execution instructions of the processor 210, and the memory 220 can be implemented by any type of volatile or non-volatile storage terminal or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. When the execution instructions in the memory 220 are executed by the processor 210, the terminal 200 can perform some or all of the steps in the following method embodiments.

[0049] The processor 210 is the control center of the storage terminal, and uses various interfaces and lines to connect various parts of the entire electronic terminal. It runs or executes software programs and / or modules stored in the memory 220, and calls data stored in the memory to perform various functions of the electronic terminal and / or process data. The processor can be composed of an integrated circuit (IC), for example, it can be composed of a single packaged IC, or it can be composed of multiple packaged ICs with the same or different functions. For example, the processor 210 can only include a central processing unit (CPU). In the embodiment of the present invention, the CPU can be a single computing core or multiple computing cores.

[0050] The communication unit 230 is used to establish a communication channel so that the storage terminal can communicate with other terminals, receive user data sent by other terminals or send user data to other terminals.

[0051] The present invention also provides a computer storage medium, wherein the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM).

[0052] The computer storage medium stores a semi-empirical and semi-physical vehicle longitudinal model building program, and the semi-empirical and semi-physical vehicle longitudinal model building program implements the following steps when executed by the processor: Based on Amesim physical model modeling software, it includes the following steps: Building a longitudinal physical model of the whole vehicle, which includes multiple sub-physical models; Simulate the longitudinal physical model of the whole vehicle, and select the sub-physical models whose number of high-frequency intervals is greater than the threshold value, and record them as the sub-physical models to be processed; Constructing a neural network model, and selecting target parameters of the sub-physical model to be processed as input and output quantities of the neural network model; Select target parameters under several working conditions of the longitudinal physical model simulation of the whole vehicle to construct a sample set to train the neural network model, and the trained neural network model is the empirical model of the sub-physical model to be processed; The obtained empirical model is imported into the original longitudinal physical model of the whole vehicle to replace the original sub-physical model to be processed, so as to generate a semi-empirical and semi-physical longitudinal model of the whole vehicle.

[0053] Those skilled in the art can clearly understand that the technology in the embodiments of the present invention can be implemented by means of software plus a necessary general hardware platform. Based on this understanding, the technical solution in the embodiments of the present invention, in essence or in other words, the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, and other media that can store program codes, including several instructions for enabling a computer terminal (which can be a personal computer, a server, or a second terminal, a network terminal, etc.) to execute all or part of the steps of the methods described in each embodiment of the present invention.

[0054] In the several embodiments provided by the present invention, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the options is only a logical function division. There may be other division methods in actual implementation, such as multiple options 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 an indirect coupling or communication connection through some interfaces, devices or options, which can be electrical, mechanical or other forms.

[0055] The options described as separate components may or may not be physically separated, and the components displayed as options may or may not be physical options, that is, they may be located in one place or distributed on multiple network options. Some or all of the options may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0056] In addition, each functional option in each embodiment of the present invention may be integrated into one processing option, or each option may exist physically separately, or two or more options may be integrated into one option.

[0057] The above disclosure is only a preferred embodiment of the present invention, but the present invention is not limited thereto. Any non-creative changes that can be thought of by a person skilled in the art, as well as several improvements and modifications made without departing from the principle of the present invention, should fall within the protection scope of the present invention.

Claims

1. A semi-empirical and semi-physical method for building a longitudinal model of a whole vehicle, characterized in that: The method is implemented based on Amesim physical model modeling software and includes the following steps: Building a longitudinal physical model of the whole vehicle, which includes multiple sub-physical models; Simulate the longitudinal physical model of the whole vehicle, and select the sub-physical models whose number of high-frequency intervals is greater than the threshold, and record them as the sub-physical models to be processed; Constructing a neural network model, and selecting target parameters of the sub-physical model to be processed as input and output quantities of the neural network model; Select target parameters under several working conditions of the longitudinal physical model simulation of the whole vehicle to construct a sample set to train the neural network model, and the trained neural network model is the empirical model of the sub-physical model to be processed; The obtained empirical model is imported into the original longitudinal physical model of the whole vehicle to replace the original sub-physical model to be processed, so as to generate a semi-empirical and semi-physical longitudinal model of the whole vehicle.

2. The semi-empirical and semi-physical vehicle longitudinal model building method according to claim 1 is characterized in that: When simulating the longitudinal physical model of the whole vehicle, the simulation type is parallel operation, the simulation solution method is variable step size solution, and four simulation conditions are used, one of which is a condition made from data collected from the actual vehicle, and the other three are standard conditions.

3. The semi-empirical and semi-physical vehicle longitudinal model building method according to claim 1 is characterized in that: Filter out sub-physical models whose number of high-frequency intervals is greater than the threshold, including: Select the frequency option in the performance analysis options, click the run button to display the high-frequency time interface, and select the sub-physical model whose number of high-frequency intervals is greater than the threshold through the high-frequency interval information on the high-frequency time interface.

4. The semi-empirical and semi-physical vehicle longitudinal model building method according to claim 1 is characterized in that: The screened sub-physical model to be processed is the engine module.

5. The semi-empirical and semi-physical vehicle longitudinal model building method according to claim 4 is characterized in that: The target parameters of the sub-physical model to be processed include the engine acceleration signal and the engine speed fed back to the engine by the engine output shaft as the input of the neural network model, and the engine output torque, engine speed, engine brake mean effective pressure, engine maximum torque and engine minimum torque as the output of the neural network model.

6. The semi-empirical and semi-physical vehicle longitudinal model building method according to any one of claims 1 to 5, characterized in that: The constructed neural network model includes two hidden layers, the first layer is a recurrent neural network, the second layer is a linear layer, the number of cell units in both layers is 9 to 11, and the number of model iterations is less than 400.

7. The semi-empirical and semi-physical vehicle longitudinal model building method according to any one of claims 1 to 5, characterized in that: The obtained empirical model is imported into the original vehicle longitudinal physical model to replace the original sub-physical model to be processed, and a semi-empirical and semi-physical vehicle longitudinal model is generated, which specifically includes: After the neural network model training is completed, the FMU file is generated in the ROM creation program; Delete the sub-physical model to be processed in the original longitudinal physical model of the whole vehicle; In the FMU import assistant option, select the FMU file generated in the ROM creation program. In the module library, drag the neural network model into the view, connect the input and output to complete the replacement of the sub-physical model to be processed, and generate a semi-empirical and semi-physical vehicle longitudinal model.

8. The semi-empirical and semi-physical vehicle longitudinal model building method according to claim 7 is characterized in that: After generating the semi-empirical and semi-physical vehicle longitudinal model, the following steps are also included: The semi-empirical and semi-physical vehicle longitudinal model is simulated, and the simulation solution type is fixed step solution, and the simulation step is 0.001s; In the Create Interface Block option, select the module type as FMI protocol and connect the output of the vehicle longitudinal physical model to the Create Interface Block option; Click the FMU Import Assistant option and select the export type to generate an FMU file of the semi-empirical and semi-physical vehicle longitudinal model.

9. A terminal, characterized in that: include: A memory for storing a semi-empirical and semi-physical vehicle longitudinal model building program; A processor is used to implement the steps of the semi-empirical and semi-physical vehicle longitudinal model building method as described in any one of claims 1 to 8 when executing the semi-empirical and semi-physical vehicle longitudinal model building program.

10. A computer-readable storage medium, characterized in that: The readable storage medium stores a semi-empirical and semi-physical vehicle longitudinal model building program, which, when executed by the processor, implements the steps of the semi-empirical and semi-physical vehicle longitudinal model building method as described in any one of claims 1 to 8.