Cloud monitoring platform and monitoring method for transient performance of air spring

Abstract

The invention provides a cloud monitoring platform and method for transient performance of an air spring, and the platform comprises an obtaining module which is used for obtaining the working condition information of the air spring; the monitoring module is used for determining the transient transmission force of the air spring according to the working condition information of the air spring and a preset hybrid model; the hybrid model comprises a network model and a physical model; the network model is used for predicting physical parameters in the calculation process of the transient transmission force of the air spring based on the working condition information of the air spring; the physical model is used for determining the transient transmission force of the air spring based on the physical parameters and the working condition information of the air spring. According to the method provided by the embodiment of the invention, the transient performance of the air spring can be quickly and accurately monitored.

Description

Technical Field

[0001] The present invention relates to the technical field of air springs, and in particular to a cloud monitoring platform and a monitoring method for transient performance of air springs. Background Art

[0002] As an important part of the vehicle suspension system, the main function of air springs is to absorb and alleviate the impact transmitted from the road surface to the vehicle body through the compressibility of air, thereby improving the ride comfort and handling stability of the vehicle. Compared with traditional steel springs, air springs have better adjustment capabilities and softer dynamic response, so they are widely used in high-end vehicles, commercial vehicles and special-purpose vehicles (such as off-road vehicles, heavy trucks, etc.).

[0003] In the related art, the performance of air springs is mostly monitored and predicted by traditional physical models. However, since traditional physical models are usually established under idealized assumptions, they cannot accurately monitor and predict the transient performance of air springs. Summary of the invention

[0004] The present invention provides a cloud monitoring platform and a monitoring method for the transient performance of an air spring, which realize rapid and accurate monitoring of the transient performance of the air spring.

[0005] The present invention provides a cloud monitoring platform for transient performance of air springs, comprising: An acquisition module, used to obtain the working condition information of the air spring; A monitoring module is used to determine the transient transmission force of the air spring based on the working condition information of the air spring and a preset hybrid model; the hybrid model includes a network model and a physical model; the network model is used to predict the physical parameters in the calculation process of the transient transmission force of the air spring based on the working condition information of the air spring; the physical model is used to determine the transient transmission force of the air spring based on the physical parameters and the working condition information of the air spring.

[0006] According to a cloud monitoring platform for transient performance of air springs provided by the present invention, the physical parameters predicted by the network model based on the working condition information of the air spring include the heat exchange coefficient of the air spring and the equivalent characteristic amplitude of the air spring; the physical model is specifically used for: Determining a transient pressure of the air spring based on a heat exchange coefficient of the air spring predicted by the network model and operating condition information of the air spring; Determining the dynamic stiffness of the rubber airbag of the air spring based on the equivalent characteristic amplitude of the air spring predicted by the network model; determining the correction term of the transient pressure of the air spring based on the dynamic stiffness of the rubber airbag of the air spring; the correction term of the transient pressure of the air spring is used to correct the transient pressure of the air spring based on the dynamic stiffness of the rubber airbag of the air spring; Based on the transient pressure of the air spring and a correction term of the transient pressure of the air spring, a transient transmission force of the air spring is determined.

[0007] According to a cloud monitoring platform for transient performance of air springs provided by the present invention, the physical model includes:

[0008] in, Indicates the transient transmission force of the air spring; Indicates the transient pressure of the air spring; A correction term representing the transient pressure of the air spring; Indicates the pressure of the atmospheric environment; Indicates the area of ​​the air spring.

[0009] According to a cloud monitoring platform for transient performance of air springs provided by the present invention, the correction term of the transient pressure of the air spring is It is determined based on the following: ； in, A correction term representing the transient pressure of the air spring; Indicates the dynamic stiffness of the rubber airbag of the air spring; represents the specific heat ratio; Indicates the transient pressure of the air spring; Indicates the area of ​​the air spring; Indicates the gas volume in the air spring; The dynamic stiffness of the rubber airbag of the air spring is determined based on the following method: ； in, Indicates the dynamic stiffness of the rubber airbag; It represents the storage modulus of the rubber airbag under the strain amplitude; represents the equivalent characteristic amplitude of the air spring predicted by the network model; Indicates the amplitude of the sinusoidal excitation corresponding to the operating condition information of the air spring at consecutive moments.

[0010] According to a cloud monitoring platform for transient performance of air springs provided by the present invention, the transient pressure of the air spring is determined based on the following method:

[0011] in, Indicates the transient pressure of the air spring; Indicates the volume of gas in the air spring; Indicates the heat exchange coefficient of the air spring; represents the specific heat ratio; Indicates the temperature of the atmospheric environment; Indicates the temperature of the air spring chamber; The derivative of the transient pressure of the air spring; Represents the derivative of the gas volume within the air spring.

[0012] According to a cloud monitoring platform for transient performance of air springs provided by the present invention, the network model is trained based on the following method: Inputting the working condition information of the air spring into the network model to obtain the heat exchange coefficient of the air spring and the equivalent characteristic amplitude of the air spring predicted by the network model; According to the heat exchange coefficient of the air spring, the equivalent characteristic amplitude of the air spring and the physical model, the transient pressure and the transient transmission force of the air spring output by the physical model are obtained; A trained network model is obtained according to the transient pressure, transient transmission force and loss function of the air spring output by the physical model; the loss function includes at least one of the following: A first loss function; the first loss function is used to evaluate the difference between the transient transmission force of the air spring obtained based on the physical model and the transmission force of the air spring obtained by measurement; A second loss function; the second loss function is used to evaluate the difference between the transient pressure of the air spring obtained based on the dynamic model and the internal pressure of the air spring obtained by measurement; The third loss function is used to evaluate the smoothness of the heat exchange coefficient of the air spring predicted by the network model.

[0013] The present invention also provides a method for monitoring the transient performance of an air spring, comprising the following steps: Get the working condition information of the air spring; The transient transmission force of the air spring is determined according to the working condition information of the air spring and a preset hybrid model; the hybrid model includes a network model and a physical model; the network model is used to predict the physical parameters in the calculation process of the transient transmission force of the air spring based on the working condition information of the air spring; the physical model is used to determine the transient transmission force of the air spring based on the physical parameters and the working condition information of the air spring.

[0014] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the above-mentioned method for monitoring transient performance of an air spring when executing the program.

[0015] The present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method for monitoring the transient performance of an air spring as described above is implemented.

[0016] The present invention also provides a computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for monitoring the transient performance of an air spring as described above is implemented.

[0017] The cloud monitoring platform and monitoring method for the transient performance of air springs provided by the present invention predict the physical parameters of the air spring that are difficult to measure in the calculation process of the transient transmission force of the air spring through a network model, identify physical parameters such as the equivalent heat exchange coefficient without adding sensors, and accurately identify the dynamic characteristics of the air spring during movement. Then, the physical model can quickly and accurately calculate the transmission force of the air spring based on the physical parameters predicted by the network model, thereby effectively improving the accuracy of the transient transmission force of the air spring and realizing rapid and accurate monitoring of the transient performance of the air spring. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings that are difficult to measure and use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are 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.

[0019] Figure 1 It is a schematic diagram of a cloud monitoring platform for transient performance of air springs provided by the present invention.

[0020] Figure 2 It is a schematic diagram of the working principle of the hybrid model provided by the present invention.

[0021] Figure 3 It is a schematic diagram of the equivalent amplitude of the dynamic stiffness of the rubber airbag provided by the present invention.

[0022] Figure 4 The figure is a working principle diagram of the air spring mass determination provided by the present invention.

[0023] Figure 5 It is a schematic diagram of the training process of the network model in the hybrid model provided by the present invention.

[0024] Figure 6 It is a schematic diagram of the air suspension system testing system provided by the present invention.

[0025] Figure 7-Figure 12 It is a schematic diagram of the experimental results provided by the present invention. DETAILED DESCRIPTION

[0026] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0027] Combine the following Figure 1-Figure 12 The cloud monitoring platform and monitoring method of the transient performance of air springs of the present invention are described.

[0028] In order to facilitate a clearer understanding of the technical solutions of the embodiments of the present application, some technical contents related to the embodiments of the present application are first introduced.

[0029] As an important part of the vehicle suspension system, the main function of air springs is to absorb and alleviate the impact transmitted from the road surface to the vehicle body through the compressibility of air, thereby improving the ride comfort and handling stability of the vehicle. Compared with traditional steel springs, air springs have better adjustment capabilities and softer dynamic response, so they are widely used in high-end vehicles, commercial vehicles and special-purpose vehicles (such as off-road vehicles, heavy trucks, etc.).

[0030] Traditional air spring control systems rely on preset physical models, which usually include the geometric parameters, material properties, and air compression laws of air springs. However, because existing physical models are usually based on idealized assumptions, the prediction accuracy under complex actual working conditions is often limited. For example, it is difficult for physical models to accurately describe the nonlinear behavior of air springs under different temperature, pressure, and load conditions. In addition, factors such as road disturbances and load changes to which vehicles are subjected during driving will cause changes in the dynamic characteristics of air springs. It is difficult for traditional physical models to fully capture these transient changes, resulting in delayed response of the control system and difficulty in achieving precise performance regulation.

[0031] In the related technology, data-driven models are gradually introduced into the performance prediction and control of air springs. Data-driven models use measured data during vehicle operation and machine learning algorithms to establish a prediction model for air spring performance. This type of model can overcome the shortcomings of physical models to a certain extent. In particular, when facing complex and nonlinear problems, data-driven models can usually provide higher prediction accuracy. The disadvantage of data-driven models is that they are highly dependent on data. If there is a lack of sufficient high-quality data, the reliability of their prediction results will be greatly reduced. In addition, due to the lack of physical explanation of data-driven models, their generalization ability and model stability may be challenged under certain extreme conditions. The emergence of hybrid models provides a solution that combines the advantages of physical models and data-driven models. Hybrid models use the theoretical guidance of physical models and the precise prediction ability of data-driven models. Through complementary advantages, they can achieve more accurate prediction and control of air spring performance under complex working conditions. Hybrid models can rely on the principles of physical models to ensure the reliability of models under extreme working conditions, and can also improve the accuracy and response speed of predictions through data-driven models.

[0032] At present, the market still lacks an air spring transient performance supervision platform and system that can fully integrate hybrid models and achieve real-time monitoring and control through cloud technology. Existing control systems are usually decentralized and independent, making it difficult to form synergy effects, and unable to fully utilize the advantages of hybrid models to improve the overall performance of air springs. With the continuous development of intelligent vehicles and autonomous driving technology, the requirements for air spring performance are becoming higher and higher. This urgently requires a new type of air spring performance supervision system that can monitor the status of air springs in real time under complex and dynamic working conditions, and dynamically adjust the control strategy through the application of hybrid models to achieve optimal suspension performance.

[0033] Figure 1 is a schematic diagram of a cloud monitoring platform for transient performance of air springs provided by the present invention, such as Figure 1 As shown, including the following: An acquisition module, used to obtain the working condition information of the air spring; A monitoring module is used to determine the transient transmission force of the air spring based on the working condition information of the air spring and a preset hybrid model; the hybrid model includes a network model and a physical model; the network model is used to predict the physical parameters in the calculation process of the transient transmission force of the air spring based on the working condition information of the air spring; the physical model is used to determine the transient transmission force of the air spring based on the physical parameters and the working condition information of the air spring.

[0034] Specifically, in order to achieve accurate monitoring and prediction of the transient performance of air springs, the cloud monitoring platform for the transient performance of air springs in the embodiment of the present application includes an acquisition module and a monitoring module. Among them, the acquisition module is used to obtain the working condition information of the air spring, such as the position information and speed information of the air spring. The monitoring module is used to determine the transient transmission force of the air spring based on the working condition information of the air spring and a preset hybrid model, so as to achieve accurate monitoring of the transient performance of the air spring. Optionally, the hybrid model in the embodiment of the present application includes a network model and a physical model. Among them, the network model is used to predict the physical parameters in the calculation process of the transient transmission force of the air spring based on the working condition information of the air spring; the physical model is used to determine the transient transmission force of the air spring based on the physical parameters and the working condition information of the air spring. Optionally, some physical parameters in the calculation process of the transient transmission force of the air spring, such as the heat exchange coefficient and the equivalent characteristic amplitude of the air spring, have a great influence on the hysteresis characteristics of the air spring under different working conditions, and play an extremely critical role in the accurate modeling and control of the transient transmission force of the air spring. However, the measurement difficulty of these physical parameters is extremely great and the accuracy is low. Therefore, in the embodiment of the present application, the physical parameters of the air spring that are difficult to measure in the calculation process of the transient transmission force of the air spring are predicted by the network model, which effectively improves the accuracy of the transient transmission force of the air spring and realizes the rapid and accurate monitoring of the transient performance of the air spring. Optionally, the monitoring platform for the transient performance of the air spring in the embodiment of the present application can be deployed on the cloud, thereby effectively improving the efficiency of data processing and monitoring of the transient performance of the air spring. The cloud monitoring platform for the transient performance of the air spring in the embodiment of the present application is universal for different air springs. The current transmission force can be identified according to the transient working condition of the air spring, and the transient monitoring and safety supervision of the dynamic stiffness of the air spring are realized, which provides a good foundation for the subsequent precise control and supervision of the suspension and the whole vehicle simulation and experimental verification.

[0035] For example, Figure 2 As shown, the working condition information of the air spring (such as the displacement information z and pressure information p of the air spring) is preprocessed and input into the network model to obtain the heat exchange coefficient and the equivalent characteristic amplitude of the air spring; then the physical model can quickly and accurately calculate the transmission force of the air spring according to the heat exchange coefficient and the equivalent characteristic amplitude of the air spring output by the network model. Optionally, the preprocessing of the working condition information may include normalization and vectorization. Optionally, the preprocessed working condition information of the air spring is input into a feedforward neural network (FFNN) and a long short-term memory network (LSTM), and then connected to each other through a fully connected network (FC), and the heat exchange coefficient and the equivalent characteristic amplitude of the air spring at different times are output and fed back to the physical model, and finally the transmission force of the air spring output by the physical model is obtained.

[0036] The cloud monitoring platform for the transient performance of air springs in the above-mentioned embodiment predicts the physical parameters of the air springs that are difficult to measure during the calculation of the transient transmission force of the air springs through a network model, identifies physical parameters such as the equivalent heat exchange coefficient without adding sensors, and accurately identifies the dynamic characteristics of the air springs during movement. The physical model can then quickly and accurately calculate the transmission force of the air spring based on the physical parameters predicted by the network model, effectively improving the accuracy of the transient transmission force of the air springs and realizing rapid and accurate monitoring of the transient performance of the air springs.

[0037] In one embodiment, the physical parameters predicted by the network model based on the working condition information of the air spring include the heat exchange coefficient of the air spring and the equivalent characteristic amplitude of the air spring; the physical model is specifically used for: Determine the transient pressure of the air spring based on the heat exchange coefficient of the air spring predicted by the network model and the working condition information of the air spring; Based on the equivalent characteristic amplitude of the air spring predicted by the network model, the dynamic stiffness of the rubber airbag of the air spring is determined; based on the dynamic stiffness of the rubber airbag of the air spring, the correction term of the transient pressure of the air spring is determined; the correction term of the transient pressure of the air spring is used to correct the transient pressure of the air spring based on the dynamic stiffness of the rubber airbag of the air spring; The transient transmission force of the air spring is determined based on the transient pressure of the air spring and a correction term of the transient pressure of the air spring.

[0038] Specifically, some physical parameters in the calculation process of the transient transmission force of the air spring, such as the heat exchange coefficient and the equivalent characteristic amplitude of the air spring, have a great influence on the hysteresis characteristics of the air spring under different working conditions, and play an extremely critical role in the accurate modeling and control of the transient transmission force of the air spring. However, the measurement of these physical parameters is extremely difficult and has low accuracy. Therefore, in the embodiment of the present application, the heat exchange coefficient of the air spring and the equivalent characteristic amplitude of the air spring are predicted by the network model, and then the physical model determines the air spring based on the heat exchange coefficient of the air spring predicted by the network model and the working condition information of the air spring. The transient pressure of the air spring is calculated based on the equivalent characteristic amplitude of the air spring predicted by the network model, and the dynamic stiffness of the rubber airbag of the air spring is determined; based on the dynamic stiffness of the rubber airbag of the air spring, the correction term of the transient pressure of the air spring is determined, so that in the process of calculating the transient transmission force of the air spring, not only the correlation between the transient pressure of the air spring and the transient transmission force of the air spring is considered, but also the influence of the dynamic stiffness generated by the rubber airbag on the transmission force is considered, and the transient pressure of the air spring is corrected by the dynamic stiffness of the rubber airbag, so that the transient transmission force of the air spring finally determined is more accurate.

[0039] The system of the above embodiment predicts the heat exchange coefficient of the air spring and the equivalent characteristic amplitude of the air spring through the network model, identifies physical parameters such as the equivalent heat exchange coefficient without adding sensors, accurately identifies the dynamic characteristics of the air spring during movement, and effectively improves the accuracy of the transient transmission force of the air spring. In addition, in the calculation process of the transmission force of the air spring, the present application fully considers the time-varying characteristics of the dynamic stiffness generated by the rubber airbag during the transient movement, and uses the dynamic stiffness of the rubber airbag to correct the transient pressure of the air spring, so that the transient transmission force of the air spring finally determined is more accurate.

[0040] In one embodiment, the physical model includes:

[0041] in, Indicates the transient transmission force of the air spring; Indicates the transient pressure of the air spring; A correction term representing the transient pressure of the air spring; Indicates the pressure of the atmospheric environment; Indicates the area of ​​the air spring.

[0042] Specifically, the physical model in the embodiment of the present application not only takes into account the correlation between the transient pressure of the air spring and the transient transmission force of the air spring, but also takes into account the influence of the dynamic stiffness generated by the rubber airbag on the transmission force, and uses the dynamic stiffness of the rubber airbag to correct the transient pressure of the air spring, accurately characterizing the transmission force of the air spring under different working conditions at the nth time step affected by the internal high-pressure gas and diaphragm stiffness, so that the determined transient transmission force of the air spring is more accurate.

[0043] In one embodiment, the correction term for the transient pressure of the air spring is It is determined based on the following: ； in, A correction term representing the transient pressure of the air spring; Indicates the dynamic stiffness of the rubber airbag of the air spring; represents the specific heat ratio; Indicates the transient pressure of the air spring; Indicates the area of ​​the air spring; Indicates the gas volume in the air spring; The dynamic stiffness of the rubber bellows of the air spring is determined based on the following method: ； in, Indicates the dynamic stiffness of the rubber airbag; It represents the storage modulus of the rubber airbag under the strain amplitude; represents the equivalent characteristic amplitude of the air spring predicted by the network model; Indicates the amplitude of the sinusoidal excitation corresponding to the operating condition information of the air spring at consecutive moments.

[0044] Specifically, in the process of calculating the transient transmission force of the air spring, not only the correlation between the transient pressure of the air spring and the transient transmission force of the air spring is taken into account, but also the influence of the dynamic stiffness of the rubber airbag on the transmission force is taken into account, and the dynamic stiffness of the rubber airbag is used to correct the transient pressure of the air spring, so that the transient transmission force of the air spring finally determined is more accurate. Optionally, the correction term of the transient pressure of the air spring in the embodiment of the present application is It is determined based on the following: ； in, A correction term representing the transient pressure of the air spring; Indicates the dynamic stiffness of the rubber airbag of the air spring; represents the specific heat ratio; Indicates the transient pressure of the air spring; Indicates the area of ​​the air spring; Indicates the gas volume in the air spring.

[0045] Optionally, since a passenger car equipped with an air spring generally travels on a paved road with good road conditions, its excitation amplitude is relatively low. Therefore, considering the correlation of the dynamic stiffness amplitude of the air spring caused by the Payne effect KPayne of the rubber material, the dynamic stiffness of the rubber airbag of the air spring is determined based on the following method: ； in, Indicates the dynamic stiffness of the rubber airbag; It represents the storage modulus of the rubber airbag under the strain amplitude; It represents the equivalent characteristic amplitude of the air spring predicted by the network model, which is related to the rubber material and structure; Indicates the amplitude of the sinusoidal excitation corresponding to the operating condition information of the air spring at consecutive moments.

[0046] Optionally, in the embodiment of the present application, the amplitude A of the sinusoidal excitation corresponding to the working condition information of the air spring at continuous moments is determined based on the following method, so that the dynamic stiffness of the rubber airbag of the air spring in a transient state can be accurately determined: The equivalent working condition method is used to calculate the displacement and velocity of the current air spring from the equilibrium position. The information of two adjacent sampling moments is used for calculation. After being equivalent to a sinusoidal excitation function, its amplitude is used as the calculation parameter of the dynamic stiffness of the rubber diaphragm at that moment. Figure 3 , take the nth time step respectively and the n-1th time step The working position of the air spring and the relative movement speed of the upper and lower covers , the amplitude is , the sinusoidal excitation frequency is ω, and the phase difference is , then the motion constraint relationship of these two time steps can be obtained:

[0047] It is not difficult to derive the expression of the amplitude at time n through the relationship of trigonometric functions, which can be further expressed as follows using known variables:

[0048] It should be noted that the amplitude needs to be limited. The limit size is the upper limit of the general test of the air spring, which is 20mm. The excitation frequency is the average value of the calculated values ​​at two moments:

[0049] The cloud monitoring platform for the transient performance of the air spring in the above-mentioned embodiment takes into account the influence of the dynamic stiffness generated by the rubber airbag on the transmission force during the calculation of the transient transmission force of the air spring, and uses the dynamic stiffness of the rubber airbag to correct the transient pressure of the air spring, so that the transient transmission force of the air spring finally determined is more accurate.

[0050] In one embodiment, the transient pressure of the air spring is determined based on:

[0051] in, Indicates the transient pressure of the air spring; Indicates the volume of gas in the air spring; Indicates the heat exchange coefficient of the air spring; represents the specific heat ratio; Indicates the temperature of the atmospheric environment; Indicates the temperature of the air spring chamber; The derivative of the transient pressure of the air spring; Represents the derivative of the gas volume within the air spring.

[0052] Specifically, during the working process of the air spring, the first law of thermodynamics and the ideal gas state equation always hold true. Therefore, based on the first law of thermodynamics and the ideal gas state equation, this application can accurately characterize the corresponding relationship between the transient pressure of the air spring and the heat exchange coefficient of the air spring and the volume of gas in the air spring, thereby realizing the accurate determination of the transient pressure of the air spring. For example, the working principle of the air spring mass determination is as follows: Figure 4 As shown, during the working process of the air spring, the gas state equation and its total differential and the first law of thermodynamics always hold true. The parameters are defined as follows: the absolute pressure P of the air spring; the positive direction of the gas mass flow is flowing into the additional air chamber; the temperature T, gas mass m, and volume V all increase in the positive direction; the spring displacement up and down , and air spring transmission force Upward is positive; for air, the ideal gas state constant R = 287 J / (kg·K); the subscripts b, 0, and atm represent the air chamber, initial state, and external atmospheric environment, respectively; the change of each parameter is represented by d; the heat Q absorbed by the gas from the outside, the work W done by the outside on the gas, the increase in internal energy U, and A represent the effective area of ​​the air spring.

[0053] Optionally, the expressions for the air spring transmitted force, volume and volume change are as follows: (1)

[0054] Optionally, the gas state equation and its differential form always hold: (2)

[0055] Combining the relationship between the constant pressure specific heat capacity Cp and the constant volume specific heat capacity CV of the gas, we can get: (3)

[0056] The first law of thermodynamics can be written as: (4)

[0057] According to the first law of thermodynamics for the air spring chamber, the heat transmitted from the outside is proportional to the temperature difference within the time Δt, and the proportionality coefficient is Since the air chamber has rolling characteristics, the outside world will do work on the gas at this time. The increase in the internal energy of the gas in the air chamber is the change in internal energy caused by the change in gas temperature. Then the parameter expression of the thermodynamic quantity can be written: (5)

[0058] Substituting equation (5) into the first law of thermodynamics and dividing both sides by the time derivative, we obtain: (6)

[0059] According to the first law of thermodynamics and the differential form of the ideal gas state equation, eliminating the temperature derivative with time, the dynamic equation of the fixed mass working process of the single-chamber air spring can be obtained as follows: (7)

[0060] in, Indicates the transient pressure of the air spring; Indicates the volume of gas in the air spring; Indicates the heat exchange coefficient of the air spring; represents the specific heat ratio; Indicates the temperature of the atmospheric environment; Indicates the temperature of the air spring chamber; The derivative of the transient pressure of the air spring; Represents the derivative of the gas volume within the air spring.

[0061] It should be noted that the dynamic model of the air spring in the embodiment of the present application during the constant mass working process accurately gives the differential equation relationship between the internal pressure of the air spring chamber and the motion change, and accurately describes the changing relationship of the internal pressure during the movement of the air spring, so that the transient pressure of the air spring can be accurately determined.

[0062] Furthermore, by using the ideal gas state equation (2) to replace the temperature term in the formula and rearranging it, the mapping relationship between the internal pressure and volume of the air spring and their changes can be obtained, thereby simplifying the transient pressure of the air spring The calculation process can realize fast and accurate monitoring of the transient performance of air springs.

[0063] (8)

[0064] The cloud monitoring platform for the transient performance of air springs in the above-mentioned embodiment accurately gives the differential equation relationship between the internal pressure of the air spring chamber and the motion change, and accurately describes the changing relationship of the internal pressure during the movement of the air spring, so that the transient pressure of the air spring can be accurately determined, thereby realizing rapid and accurate monitoring of the transient performance of the air spring.

[0065] In one embodiment, the network model is trained based on the following method: Inputting the working condition information of the air spring into the network model, obtaining the heat exchange coefficient of the air spring and the equivalent characteristic amplitude of the air spring predicted by the network model; According to the heat exchange coefficient of the air spring, the equivalent characteristic amplitude of the air spring and the physical model, the transient pressure and transient transmission force of the air spring output by the physical model are obtained; According to the transient pressure, transient transmission force and loss function of the air spring output by the physical model, a trained network model is obtained; the loss function includes at least one of the following: A first loss function; the first loss function is used to evaluate the difference between the transient transmission force of the air spring obtained based on the physical model and the transmission force of the air spring obtained by measurement; A second loss function; the second loss function is used to evaluate the difference between the transient pressure of the air spring obtained based on the dynamic model and the internal pressure of the air spring obtained by measurement; The third loss function is used to evaluate the smoothness of the heat exchange coefficient of the air spring predicted by the network model.

[0066] Specifically, the determination of the transient pressure of the air spring and the correction term of the transient pressure of the air spring involves the determination of physical parameters such as the heat exchange coefficient and the equivalent characteristic amplitude of the air spring, and these physical parameters have a great influence on the hysteresis characteristics of the air spring under different working conditions, and play an extremely critical role in the accurate modeling and control of the transient transmission force of the air spring. However, these physical parameters are difficult to measure and express accurately in actual engineering. Therefore, the present application predicts the physical parameters of the air spring that are difficult to measure in the calculation process of the transient transmission force of the air spring through a network model, thereby effectively improving the accuracy of the transient transmission force of the air spring and realizing rapid and accurate monitoring of the transient performance of the air spring.

[0067] Optionally, the network model in the embodiment of the present application is trained based on the following method: First, the working condition information of the air spring is input into the network model to obtain the heat exchange coefficient of the air spring and the equivalent characteristic amplitude of the air spring predicted by the network model; optionally, the working condition information of the air spring includes the position information and speed information of the air spring. Then, according to the heat exchange coefficient of the air spring, the equivalent characteristic amplitude of the air spring and the physical model, the transient pressure and transient transmission force of the air spring output by the physical model are obtained. Finally, according to the transient pressure, transient transmission force and loss function of the air spring output by the physical model, the trained network model is obtained. Optionally, the loss function is used to measure the error between the transient pressure and transient transmission force of the air spring output by the physical model and the actual measurement value, so as to realize the training of the network model.

[0068] Exemplarily, the training process of the network model in the hybrid model in the embodiment of the present application is as follows: Figure 5As shown, in the process of predicting the heat exchange coefficient based on the network model, since the equivalent heat exchange coefficient between the air spring and the outside world is not only related to the current moment, but also related to the loading history and cannot change suddenly, the working condition information at continuous moments (such as the displacement information z and pressure information p of the air spring) is pre-processed and input into the network model in this application to obtain the heat exchange coefficient and the equivalent characteristic amplitude of the air spring; then the physical model can calculate the transient pressure and transient transfer force of the air spring according to the heat exchange coefficient and the equivalent characteristic amplitude of the air spring output by the network model, and further train the network model based on the calculation results and loss function of the physical model, so that the trained network model can output more accurate physical parameters, thereby effectively improving the accuracy of the transfer force output by the physical model. Optionally, the preprocessing of the working condition information can include normalization processing and vectorization processing, etc. Optionally, such as Figure 5 As shown, the preprocessed working condition information of the air spring is input into the feedforward neural network (FFNN) and the long short-term memory network (LSTM), and then connected to each other through the fully connected network (FC). The heat exchange coefficient and the equivalent characteristic amplitude of the air spring at different times are output and fed back to the physical model, and the loss function is calculated to train the network model.

[0069] Optionally, the loss function in the embodiment of the present application includes at least one of the following: A first loss function; the first loss function is used to evaluate the difference between the transient transmission force of the air spring obtained based on the physical model and the transmission force of the air spring obtained by measurement; A second loss function; the second loss function is used to evaluate the difference between the transient pressure of the air spring obtained based on the dynamic model and the internal pressure of the air spring obtained by measurement; The third loss function is used to evaluate the smoothness of the heat exchange coefficient of the air spring predicted by the network model.

[0070] Specifically, the error calculation of the loss function in the embodiment of the present application is divided into three parts according to the actual physical meaning.

[0071] In the first part, the theoretical value of the force is calculated according to the modified air spring force transmission formula and compared with the actual sensor measurement value to calculate the mean square error (MSE).

[0072] In the second part, the MSE is calculated using the internal pressure recursively obtained from the physical model and the measured pressure.

[0073] In the third part, considering that the output of the neural network is the heat transfer coefficient, which should change continuously and with a limited amplitude, the smoothness penalty (SP) term is selected as part of the loss function, and λ is the adjustment factor.

[0074] Alternatively, the overall loss function can be written as:

[0075] That is, during the training process of the network model, the present application not only considers the error between the transient transmission force of the air spring obtained by the physical model and the transmission force of the air spring obtained by measurement, but also considers the error between the transient pressure of the air spring obtained based on the dynamic model and the internal pressure of the air spring obtained by measurement, and also considers the smoothness of the heat exchange coefficient of the air spring predicted by the network model. As a result, the trained network model can more accurately predict the physical parameters of the air spring that are difficult to measure in the process of calculating the transient transmission force of the air spring, effectively improve the accuracy of the transient transmission force of the air spring, and realize the rapid and accurate monitoring of the transient performance of the air spring.

[0076] For example, Figure 6 As shown, in order to test the accuracy of the transient transmission force of the air spring determined by the cloud monitoring platform of the transient performance of the air spring in the embodiment of the present application, a test bench for the air suspension system was built in the embodiment of the present application, and the SUV rear axle air spring was selected as the test platform, and the MTS (Mechanical Testing & Simulation) dynamometer test bench and pneumatic system were designed. The pneumatic system includes a connecting pipe, a high-pressure air source, a pressure reducing valve F, a pressure sensor P connected to the air spring, etc. The pressure sensor has a range of 0 to 3MPa, and its data is read by a signal receiving unit. The hydraulic actuator generates the designed displacement excitation, and the displacement and force values ​​are read by the MTS system, and the sampling frequency of all data is 200Hz. The air spring parameters and environment-related constants during the test are shown in Table 1: Table 1

[0077] Tests were conducted on the above-mentioned test bench, and the physical model, data-driven method and the cloud monitoring platform for transient performance of air springs in this application were selected for comparison. Figure 7 and Figure 8The comparison results at 5 mm / 1 Hz and 20 mm / 0.5 Hz are shown respectively; the Data-driven curve in the figure represents the transient pressure and transmission force of the air spring determined based on the traditional data-driven method; the Physics curve in the figure represents the transient pressure and transmission force of the air spring determined based on the traditional physical model; the Original curve in the figure represents the measured pressure and transmission force of the air spring; the Hybrid curve in the figure represents the pressure and transmission force of the air spring determined based on the hybrid model in this application. It can be seen that the cloud monitoring platform for the transient performance of air springs in this application has the characteristics of high accuracy and strong universality. At the same time, the changing trend of the intermediate variables can also be detected based on the output results, such as selecting the heat exchange coefficient and rubber airbag stiffness correction items that are not easy to measure directly. Fig. 9 Shows the heat transfer coefficient under two conditions and the correction term for the transient pressure of the air spring Change trend; the blue curve represents the change trend of the heat exchange coefficient of the air spring; the yellow curve represents the correction term of the transient pressure of the air spring Changing trends. Fig.10 and Fig.11 The change trends of the two intermediate variables within the range of the experiment are shown respectively. It can be seen that the heat exchange coefficient It is highly correlated with the speed of movement and exhibits a nearly normal distribution at a single position. In a physical sense, when the absolute value of the speed is greater than a certain value, the entire internal gas change process can be approximately regarded as an adiabatic process, and no heat exchange occurs between the inside and the outside. The correction term for the transient pressure of the air spring It shows a linear relationship with speed and is greater than 1. From a physical point of view, it means that the correction term of the rubber airbag to the total stiffness cannot be ignored. As the compression speed increases, the Payne effect of the rubber material becomes stronger. Fig.12 Shows the heat transfer coefficient The fitting results are obtained by slicing at different motion positions and assuming that the mean, variance and peak value of the normal distribution show a three-fold change law with the motion position. The total error of this application can be kept below 2.95% (for pressure prediction error) and below 1.76% for transmission force under different working conditions.

[0078] It can be seen from the above experiments that the cloud monitoring platform for the transient performance of air springs in the embodiment of the present application combines the physical model with the network model, and realizes comprehensive supervision of the transient performance of air springs through cloud data processing and remote control technology. This can not only significantly improve the dynamic response capability and control accuracy of the air spring, and improve its response capability and stability under complex working conditions, but also realize adaptive adjustment of the vehicle suspension system under different working conditions through data feedback and continuous optimization, thereby improving the overall performance and driving safety of the vehicle.

[0079] Exemplarily, an embodiment of the present application further provides a method for monitoring transient performance of an air spring, which is applied to a cloud monitoring platform for transient performance of an air spring, and includes: Get the working condition information of the air spring; The transient transmission force of the air spring is determined according to the working condition information of the air spring and a preset hybrid model; the hybrid model includes a network model and a physical model; the network model is used to predict the physical parameters in the calculation process of the transient transmission force of the air spring based on the working condition information of the air spring; the physical model is used to determine the transient transmission force of the air spring based on the physical parameters and the working condition information of the air spring.

[0080] Specifically, in order to achieve accurate monitoring and prediction of the transient performance of the air spring, the embodiment of the present application first obtains the working condition information of the air spring, such as the position information and speed information of the air spring. Then, based on the working condition information of the air spring and a preset hybrid model, the transient transmission force of the air spring is determined, thereby achieving accurate monitoring of the transient performance of the air spring. Optionally, the hybrid model in the embodiment of the present application includes a network model and a physical model. Among them, the network model is used to predict the physical parameters in the calculation process of the transient transmission force of the air spring based on the working condition information of the air spring; the physical model is used to determine the transient transmission force of the air spring based on the physical parameters and the working condition information of the air spring. Optionally, some physical parameters in the calculation process of the transient transfer force of the air spring, such as the heat exchange coefficient and the equivalent characteristic amplitude of the air spring, have a great influence on the hysteresis characteristics of the air spring under different working conditions, and play an extremely critical role in the accurate modeling and control of the transient transfer force of the air spring. However, the measurement of these physical parameters is extremely difficult and has low accuracy. Therefore, in the embodiment of the present application, the physical parameters of the air spring that are difficult to measure in the calculation process of the transient transfer force of the air spring are predicted through a network model, which effectively improves the accuracy of the transient transfer force of the air spring and realizes the rapid and accurate monitoring of the transient performance of the air spring.

[0081] The method for monitoring the transient performance of the air spring in the above-mentioned embodiment predicts the physical parameters of the air spring that are difficult to measure in the process of calculating the transient transmission force of the air spring through a network model, identifies physical parameters such as the equivalent heat exchange coefficient without adding sensors, and accurately identifies the dynamic characteristics of the air spring during movement. Then, the physical model can quickly and accurately calculate the transmission force of the air spring based on the physical parameters predicted by the network model, thereby effectively improving the accuracy of the transient transmission force of the air spring and realizing rapid and accurate monitoring of the transient performance of the air spring.

[0082] The present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the monitoring method of the transient performance of the air spring provided by the above-mentioned methods, and the method includes: obtaining the working condition information of the air spring; determining the transient transmission force of the air spring according to the working condition information of the air spring and a preset hybrid model; the hybrid model includes a network model and a physical model; the network model is used to predict the physical parameters in the calculation process of the transient transmission force of the air spring based on the working condition information of the air spring; the physical model is used to determine the transient transmission force of the air spring based on the physical parameters and the working condition information of the air spring.

[0083] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, it is implemented to execute the method for monitoring the transient performance of an air spring provided by the above-mentioned methods, the method comprising: obtaining the working condition information of the air spring; determining the transient transmission force of the air spring according to the working condition information of the air spring and a preset hybrid model; the hybrid model comprises a network model and a physical model; the network model is used to predict the physical parameters in the calculation process of the transient transmission force of the air spring based on the working condition information of the air spring; the physical model is used to determine the transient transmission force of the air spring based on the physical parameters and the working condition information of the air spring.

[0084] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0085] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0086] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A cloud monitoring platform for transient performance of air springs, characterized in that: include: An acquisition module, used to obtain the working condition information of the air spring; A monitoring module, used to determine the transient transmission force of the air spring according to the working condition information of the air spring and a preset hybrid model; the hybrid model includes a network model and a physical model; the network model is used to predict the physical parameters in the calculation process of the transient transmission force of the air spring based on the working condition information of the air spring; The physical model is used to determine the transient transmission force of the air spring based on the physical parameters and the working condition information of the air spring.

2. The cloud monitoring platform for transient performance of air springs according to claim 1 is characterized in that: The physical parameters predicted by the network model based on the working condition information of the air spring include the heat exchange coefficient of the air spring and the equivalent characteristic amplitude of the air spring; the physical model is specifically used for: Determining a transient pressure of the air spring based on a heat exchange coefficient of the air spring predicted by the network model; Determining the dynamic stiffness of the rubber airbag of the air spring based on the equivalent characteristic amplitude of the air spring predicted by the network model; Determining a correction term for the transient pressure of the air spring based on the dynamic stiffness of the rubber airbag of the air spring; The correction term of the transient pressure of the air spring is used to correct the transient pressure of the air spring based on the dynamic stiffness of the rubber airbag of the air spring; Based on the transient pressure of the air spring and a correction term of the transient pressure of the air spring, a transient transmission force of the air spring is determined.

3. The cloud monitoring platform for transient performance of air springs according to claim 2 is characterized in that: The physical model includes: in, Indicates the transient transmission force of the air spring; Indicates the transient pressure of the air spring; A correction term representing the transient pressure of the air spring; Indicates the pressure of the atmospheric environment; Indicates the area of ​​the air spring.

4. The cloud monitoring platform for transient performance of air springs according to claim 3 is characterized in that: The correction term for the transient pressure of the air spring is It is determined based on the following: ； in, A correction term representing the transient pressure of the air spring; Indicates the dynamic stiffness of the rubber airbag of the air spring; represents the specific heat ratio; Indicates the transient pressure of the air spring; Indicates the area of ​​the air spring; Indicates the gas volume in the air spring; The dynamic stiffness of the rubber airbag of the air spring is determined based on the following method: ； in, Indicates the dynamic stiffness of the rubber airbag; It represents the storage modulus of the rubber airbag under the strain amplitude; represents the equivalent characteristic amplitude of the air spring predicted by the network model; Indicates the amplitude of the sinusoidal excitation corresponding to the operating condition information of the air spring at consecutive moments.

5. The cloud monitoring platform for transient performance of air springs according to claim 4 is characterized in that: The transient pressure of the air spring is determined based on the following method: in, Indicates the transient pressure of the air spring; Indicates the volume of gas in the air spring; Indicates the heat exchange coefficient of the air spring; represents the specific heat ratio; Indicates the temperature of the atmospheric environment; Indicates the temperature of the air spring chamber; The derivative of the transient pressure of the air spring; Represents the derivative of the gas volume within the air spring.

6. The cloud monitoring platform for transient performance of air springs according to any one of claims 1 to 5, characterized in that: The network model is trained based on the following method: Inputting the working condition information of the air spring into the network model to obtain the heat exchange coefficient of the air spring and the equivalent characteristic amplitude of the air spring predicted by the network model; According to the heat exchange coefficient of the air spring, the equivalent characteristic amplitude of the air spring and the physical model, the transient pressure and the transient transmission force of the air spring output by the physical model are obtained; A trained network model is obtained according to the transient pressure, transient transmission force and loss function of the air spring output by the physical model; the loss function includes at least one of the following: A first loss function; the first loss function is used to evaluate the difference between the transient transmission force of the air spring obtained based on the physical model and the transmission force of the air spring obtained by measurement; A second loss function; the second loss function is used to evaluate the difference between the transient pressure of the air spring obtained based on the dynamic model and the internal pressure of the air spring obtained by measurement; The third loss function is used to evaluate the smoothness of the heat exchange coefficient of the air spring predicted by the network model.

7. A method for monitoring transient performance of an air spring, characterized in that: A cloud monitoring platform for transient performance of an air spring as claimed in any one of claims 1 to 6, comprising: Get the working condition information of the air spring; The transient transmission force of the air spring is determined according to the working condition information of the air spring and a preset hybrid model; the hybrid model includes a network model and a physical model; the network model is used to predict the physical parameters in the calculation process of the transient transmission force of the air spring based on the working condition information of the air spring; the physical model is used to determine the transient transmission force of the air spring based on the physical parameters and the working condition information of the air spring.

8. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for monitoring transient performance of an air spring as claimed in claim 7 is implemented.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the method for monitoring transient performance of an air spring as claimed in claim 7 is implemented.

Images