Electronic water pump multi-mode cooperative control method and system based on LIN / PWM communication, electronic equipment and storage medium

Abstract

The invention provides an electronic water pump multi-mode cooperative control method and system based on LIN / PWM communication, electronic equipment and a storage medium, and relates to the technical field of vehicle thermal management systems.The method includes the steps that a vehicle turbocharging system pipeline pressure signal is collected, and frequency band energy analysis is conducted to obtain frequency spectrum energy information; a laser beam is emitted by the surface of an intercooler of a helium-neon laser, and vibration signals are inverted to obtain vibration mode information; the conditional generative adversarial network establishes a pressure and vibration incidence relation and generates a system coupling confidence parameter; multi-water pump working point distributed optimization is carried out to generate a cooperative control instruction; constructing a PWM signal sequence to form a control signal; and finally, a control signal is transmitted in a time division multiple access mode based on the LIN / PWM communication bus, and multi-mode cooperative control of the electronic water pump is achieved. Intelligent cooperative control of multiple electronic water pumps in a vehicle thermal management system can be realized.

Description

Technical Field

[0001] This application relates to the field of vehicle thermal management system technology, and in particular to a multi-mode collaborative control method, system, electronic device and storage medium for an electronic water pump based on LIN / PWM communication. Background Technology

[0002] In turbocharged systems, regulating intercooler temperature fluctuations is a key technical challenge. When engine operating conditions change drastically, the temperature of the boosted air fluctuates rapidly, requiring timely adjustments to the cooling system's operation to prevent thermal runaway. This scenario demands that the control system quickly identify temperature change trends and coordinate the operating modes of multiple electric water pumps to ensure heat dissipation efficiency and system stability.

[0003] Existing solutions typically employ model predictive control (MMC) to predict temperature changes by establishing a thermodynamic model of the intercooler and then using a closed-loop control algorithm to adjust the pump speed. This method collects temperature data from sensors, combines it with a pre-established heat transfer model for calculations, and ultimately generates corresponding control commands.

[0004] However, this approach has significant limitations. Its control accuracy heavily relies on the model's accuracy, while in real-world vehicle operating environments, complex and variable conditions make it difficult for the model to comprehensively cover all operating conditions. Furthermore, this approach exhibits a lag in responding to sudden temperature changes, failing to effectively address the instantaneous thermal shocks generated by the turbocharger system. This may result in untimely temperature regulation, impacting the overall system performance. Summary of the Invention

[0005] The purpose of this application is to provide a multi-mode cooperative control method, system, electronic device and storage medium for electronic water pumps based on LIN / PWM communication, so as to solve the problems of lag control response and low cooperative efficiency of multiple water pumps when the cooler temperature changes suddenly in the turbocharger system in the prior art.

[0006] To address the aforementioned technical problems, in a first aspect, this application provides a multi-mode cooperative control method for electronic water pumps based on LIN / PWM communication, comprising: Based on the pipeline pressure signal of the vehicle turbocharger system, the pressure pulsation is analyzed by frequency band energy using fast Fourier transform to generate spectral energy information. A laser beam is emitted onto the surface of the vehicle intercooler using a helium-neon laser, and vibration mode information is generated by inverting the vibration signal of the vehicle intercooler surface based on the changes in interference fringes. Based on the spectral energy information and the vibration mode information, a conditional generative adversarial network is used to establish the correlation between vibration and pressure pulsation, and system coupling confidence parameters are generated. Based on the system coupling confidence parameters, the alternating direction multiplier method is used to perform distributed optimization of the operating points of multiple electronic water pumps and generate cooperative control commands. Based on the aforementioned collaborative control instructions, a PWM signal sequence is constructed using pulse width modulation (PWM) technology to generate a control signal. Based on the LIN / PWM communication bus, the control signals are transmitted using time-division multiple access to generate multi-mode coordinated control signals for the electronic water pump.

[0007] Optionally, based on the spectral energy information and the vibration mode information, a conditional generative adversarial network is used to establish the correlation between vibration and pressure pulsation, generating system coupling confidence parameters, including: The spectral energy information is subjected to band energy recombination processing to obtain recombined energy data; The vibration mode information is subjected to mode shape node analysis to obtain analytical feature data; The recombined energy data is processed using a conditional generation network to perform vibration mode simulation, resulting in simulated mode data. The simulated modal data and the analytical feature data are input into a discriminant network for distribution consistency discrimination processing to obtain a distribution difference index. The distribution difference index is subjected to confidence level transformation to obtain preliminary confidence parameters; Based on the operating data of the vehicle turbocharger system, the preliminary confidence parameters are dynamically corrected to generate system coupling confidence parameters.

[0008] Optionally, a laser beam is emitted onto the surface of the vehicle intercooler using a helium-neon laser, and the vibration signal of the vehicle intercooler surface is inverted based on the changes in interference fringes to generate vibration mode information, including: Using the helium-neon laser, a laser beam is projected onto the surface of the vehicle's intercooler to obtain a reflected laser signal; The reflected laser signal is subjected to interference fringe generation processing to obtain a pattern of alternating bright and dark fringes; The light and dark alternating stripe pattern is continuously acquired and processed to obtain a stripe dynamic sequence; Based on the variation law of stripe spacing, the stripe dynamic sequence is subjected to spacing quantization processing to obtain stripe spacing data; Based on the preset laser wavelength parameters, the stripe spacing data is processed by displacement conversion to obtain the surface micro-displacement. The surface micro-displacement is modally decomposed using a mode reconstruction algorithm to obtain vibration mode information.

[0009] Optionally, based on the system coupling confidence parameters, the alternating direction multiplier method is used to perform distributed optimization of the operating points of multiple electronic water pumps, generating cooperative control commands, including: The system coupling confidence parameters are decomposed using multi-pump control constraints to obtain a distributed constraint set; An initial working point is set for the distributed constraint set to obtain an initial working state set; Parallel optimization computation is performed on the initial working state set to obtain a local optimization result set; The local optimization result set is processed by exchanging intermediate variables through the communication network of the LIN / PWM communication bus to obtain global coordination variables; The global coordination variables are updated iteratively at work points to obtain an updated work state set; Based on preset convergence criteria, the updated working state set is judged and processed to generate collaborative control instructions.

[0010] Optionally, the recombined energy data is processed using a conditional generation network to perform vibration mode simulation to obtain simulated mode data, including: Perform a pressure feature encoding operation on the recombined energy data to obtain a pressure feature vector; Perform a latent space transformation operation on the pressure feature vector to obtain a latent feature representation; By combining the operating data of the vehicle turbocharger system, a conditional constraint operation is performed on the potential feature representation to obtain a conditional feature representation; A nonlinear transformation operation is performed on the conditional feature representation using a multi-layer perceptron architecture to obtain transformed feature data; Perform a spatial reconstruction operation on the transformed feature data to obtain preliminary modal data; Perform detail enhancement operations on the preliminary modal data to obtain simulated modal data.

[0011] Optionally, the helium-neon laser is used to project a laser beam onto the surface of the vehicle intercooler to obtain a reflected laser signal, including: A single-wavelength laser beam is generated using the helium-neon laser described above. The single-wavelength laser beam is shaped using an optical lens group to form a uniform light spot; The uniform light spot is divided into a reference beam and a measurement beam using a beam splitter prism; The measurement beam is directed to a specific measurement point on the surface of the vehicle's intercooler using a set of mirrors. Receive the scattered light beam returned from the specific measurement point to obtain the original optical signal; The original optical signal is subjected to interference processing with the reference beam to obtain the reflected laser signal.

[0012] Optionally, the local optimization result set is processed through the communication network of the LIN / PWM communication bus to obtain global coordination variables, including: Using the communication network of the LIN / PWM communication bus, the local optimization result set is encapsulated to obtain a standardized data packet; Based on the time division multiple access communication protocol, the standardized data packets are processed by time slot allocation to obtain a scheduled transmission sequence; Through the physical link of the communication network, the scheduled transmission sequence is processed into data packets to obtain acknowledgment signals returned by each distributed node; The system receives confirmation signals from each distributed node, performs transmission integrity verification, and obtains data packets from each node. The data packets of each node are processed by protocol parsing to extract intermediate variable data; The intermediate variable data is processed collaboratively using a variable integration algorithm to generate a globally coordinated variable.

[0013] Secondly, this application provides a multi-mode cooperative control system for an electronic water pump based on LIN / PWM communication, including: The pressure spectrum module is used to perform frequency band energy analysis on the pipeline pressure signal based on the vehicle turbocharger system, and generate spectrum energy information by using fast Fourier transform to analyze the pressure pulsation. The laser vibration measurement module is used to emit a laser beam to the surface of the vehicle intercooler using a helium-neon laser, and invert the vibration signal of the vehicle intercooler surface based on the changes in interference fringes to generate vibration mode information; The coupling confidence module is used to establish the correlation between vibration and pressure pulsation based on the spectral energy information and the vibration mode information, combined with a conditional generative adversarial network, and to generate system coupling confidence parameters. The collaborative optimization module is used to perform distributed optimization of the operating points of multiple electronic water pumps based on the system coupling confidence parameters and using the alternating direction multiplier method to generate collaborative control commands. The signal generation module is used to construct a PWM signal sequence based on the cooperative control command using pulse width modulation technology to generate a control signal; The bus transmission module is used to transmit the control signals based on the LIN / PWM communication bus and in a time-division multiple access manner, and to generate multi-mode coordinated control signals for the electronic water pump.

[0014] Thirdly, this application provides an electronic device, comprising: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of a multi-mode cooperative control method for an electronic water pump based on LIN / PWM communication as described in the first aspect above.

[0015] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps of the multi-mode cooperative control method for an electronic water pump based on LIN / PWM communication as described in the first aspect above.

[0016] This application provides a multi-mode cooperative control method for electronic water pumps based on LIN / PWM communication. The method utilizes Fast Fourier Transform (FFT) to analyze the frequency band energy of pressure pulsations based on the pipeline pressure signal of a vehicle's turbocharger system, generating spectral energy information. A helium-neon laser beam is emitted onto the surface of the vehicle's intercooler, and vibration signal on the intercooler surface is inverted based on interference fringe changes, generating vibration mode information. Based on the spectral energy information and the vibration mode information, a conditional generative adversarial network (GAN) is used to establish the correlation between vibration and pressure pulsations, generating system coupling confidence parameters. Based on these system coupling confidence parameters, a distributed optimization of the operating points of multiple electronic water pumps is performed using the alternating direction multiplier method, generating cooperative control commands. Based on these cooperative control commands, a PWM signal sequence is constructed using pulse width modulation (PWM) technology to generate control signals. Finally, based on the LIN / PWM communication bus, the control signals are transmitted using time-division multiple access (TDMA) to generate multi-mode cooperative control signals for the electronic water pumps.

[0017] The technical solution of this application has the following beneficial effects: Based on the pipeline pressure signal of the vehicle turbocharger system, a fast Fourier transform is used to perform frequency band energy analysis on the pressure pulsation, which can accurately capture the characteristic frequency components in the pressure fluctuation and provide a basis for system state identification. A helium-neon laser is used to emit a laser beam onto the surface of the vehicle's intercooler, and the vibration signal is inverted based on the changes in interference fringes, achieving non-contact, high-precision measurement of the micro-vibrations on the intercooler surface. Based on spectral energy information and vibration mode information, a conditional generative adversarial network is used to establish the correlation between vibration and pressure pulsation, effectively identifying the intrinsic relationship between pressure and vibration. Based on the system coupling confidence parameters, the alternating direction multiplier method is used for distributed optimization of the operating points of multiple electronic water pumps, realizing the rapid solution of multi-pump collaborative control parameters. Based on the collaborative control commands, a PWM signal sequence is constructed using pulse width modulation technology, which can generate accurate water pump motor drive signals. Based on the LIN / PWM communication bus, a time-division multiple access method is used to transmit control signals, ensuring reliable transmission and synchronous execution of multi-pump control commands.

[0018] Furthermore, reconstructed energy data is obtained by performing frequency band energy reconstruction processing on the spectral energy information, and analytical feature data is obtained by performing mode shape node analysis processing on the vibration modal information. Simulated modal data is obtained by using a conditional generation network to perform vibration modal simulation processing on the reconstructed energy data. The simulated modal data and analytical feature data are then input into a discriminant network for distribution consistency discrimination processing to obtain a distribution difference index. The distribution difference index is then processed by confidence level transformation to obtain preliminary confidence parameters. Finally, the preliminary confidence parameters are dynamically corrected based on the operating condition data of the vehicle turbocharger system, ultimately generating the system coupling confidence parameters. This method achieves accurate quantification of the multi-physics coupling relationship between pressure and vibration through a multi-level processing flow. The adversarial training mechanism of generative adversarial networks improves the accuracy of coupling relationship identification, and the dynamic correction based on operating condition data ensures the adaptability of the system coupling confidence parameters under different operating conditions, providing a reliable parameter basis for subsequent distributed optimization control. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating a multi-mode collaborative control method for an electronic water pump based on LIN / PWM communication, provided in an embodiment of this application; Figure 2 A schematic diagram illustrating a specific implementation of a multi-mode collaborative control method for an electronic water pump based on LIN / PWM communication, provided in this application embodiment; Figure 3 A schematic diagram illustrating a specific implementation of a multi-mode collaborative control method for an electronic water pump based on LIN / PWM communication, provided in this application embodiment; Figure 4 This is a schematic diagram of a multi-mode collaborative control system for an electronic water pump based on LIN / PWM communication, provided as an embodiment of this application. Detailed Implementation

[0021] Research has found that in scenarios involving sudden temperature changes in the cooler of turbocharged systems, existing control schemes primarily rely on predictive control methods based on fixed thermodynamic models. This method collects data from a limited number of temperature sensors and performs calculations using a pre-established heat transfer model to adjust the cooling system's operating state. However, this approach suffers from two fundamental drawbacks: first, its control accuracy is highly dependent on the model's accuracy, and the complex and variable operating conditions in real-world vehicles make it difficult for the model to comprehensively cover all operating conditions; second, there is a significant lag in the response to sudden temperature changes, making it unable to effectively cope with the instantaneous thermal shock generated by the turbocharged system, resulting in untimely temperature regulation and impacting the overall system performance and reliability.

[0022] To address the aforementioned issues, this application proposes a multi-mode collaborative control method and system for electronic water pumps based on LIN / PWM communication. Specifically, this method first synchronously acquires pipeline pressure pulsation signals and intercooler surface vibration signals, and extracts spectral energy features and vibration mode features using Fast Fourier Transform and Laser Interferometry, respectively. Then, a conditional generative adversarial network is used to establish the coupling relationship between pressure and vibration, generating system coupling confidence parameters. Finally, a distributed optimization algorithm is employed to calculate the collaborative operating point of multiple water pumps, and precise transmission of control commands is achieved through the LIN / PWM communication bus. This solution overcomes the limitations of traditional single-temperature sensing modes, proactively capturing temperature change trends through dual-mode sensing of pressure and vibration, and utilizing intelligent algorithms to achieve collaborative optimization control of multiple water pumps. This fundamentally solves the problems of response lag and insufficient control accuracy in existing technologies, significantly improving the regulation performance and reliability of turbocharged systems under sudden temperature changes.

[0023] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] The core of this application is to provide a multi-mode cooperative control method for electronic water pumps based on LIN / PWM communication. A flowchart of one specific implementation is shown below. Figure 1 As shown, the method includes: S101. Based on the pipeline pressure signal of the vehicle turbocharger system, the pressure pulsation is analyzed by frequency band energy using fast Fourier transform to generate spectral energy information. In this step, the pipeline pressure signal refers to the amount of dynamic pressure fluctuation in the intake pipeline of the turbocharger system; The Fast Fourier Transform (FFT) is a mathematical processing method that converts a time-domain signal into a frequency-domain signal. Frequency band energy analysis refers to the process of dividing the spectrum according to frequency range and calculating the energy value of each frequency band; Spectral energy information refers to a set of data that characterizes the energy distribution features of each frequency component.

[0025] In this embodiment, the dynamic pressure signal in the pipeline is first acquired by a high-frequency pressure sensor installed at the outlet of the turbocharger. Then, the acquired time-domain pressure signal is converted into a frequency-domain representation using a fast Fourier transform algorithm. Next, the obtained spectrum signal is bandpass filtered to extract components in a specific frequency range. Then, the integral value of the signal energy in each filtered frequency band is calculated. Finally, these energy values ​​are arranged and combined in order of frequency to form complete spectrum energy information.

[0026] In a real-world scenario, when a vehicle suddenly accelerates to overtake on a highway, the turbocharger rapidly increases the boost pressure, causing pressure fluctuations in the intake manifold within a specific frequency range. Through this step, the system can accurately capture the characteristic frequency components of these pressure fluctuations and quantify them into spectral energy data suitable for subsequent analysis, providing a reliable input for vibration correlation analysis.

[0027] S102. A laser beam is emitted onto the surface of the vehicle intercooler using a helium-neon laser, and the vibration signal of the vehicle intercooler surface is inverted based on the change of interference fringes to generate vibration mode information. In this step, a helium-neon laser refers to an optical emitting device capable of emitting a laser beam of a specific wavelength; Interference fringe variation refers to the movement and shape change of bright and dark fringes in a laser interference pattern; Inversion refers to the calculation process of deriving the original state from observation results; Vibration signal refers to the measured value of the physical quantity of microscopic vibration on the surface of an object; Vibration modal information refers to the set of mode parameters that reflect the vibration characteristics of a structure.

[0028] In this embodiment, a coherent laser beam is first emitted from a helium-neon laser to a specific measurement point on the surface of the intercooler. Then, an optical receiving device is used to receive the laser reflected from the surface and cause it to interfere with the reference beam. Next, a high-speed image sensor is used to continuously capture the dynamic changes of the interference fringes. Then, the surface micro-vibration displacement data is calculated by the fringe displacement. Finally, a modal analysis algorithm is used to extract feature parameters from the vibration data to form complete vibration modal information.

[0029] Continuing with the above case, when a sudden change in the temperature of the boosted air causes the surface of the intercooler to vibrate slightly due to thermal stress, the laser interferometry technology in this step allows the system to accurately capture the micron-level vibration changes on the surface in a non-contact manner. These physical vibration signals are then converted into modal parameters with engineering application value, providing accurate data support for subsequent multiphysics coupling analysis.

[0030] S103. Based on the spectral energy information and the vibration mode information, and combined with the conditional generative adversarial network, establish the correlation between vibration and pressure pulsation, and generate system coupling confidence parameters; In this step, a conditional generative adversarial network refers to a deep learning architecture that includes a generator and a discriminator; The relationship of correlation refers to the mutual influence and constraint between different physical quantities; The system coupling confidence parameter is a reliability index that quantifies the degree of coupling between pressure and vibration.

[0031] In this embodiment, the spectral energy information is first input into the generator network as a condition to generate simulated vibration data. Then, the real vibration mode information and the generated simulated data are simultaneously input into the discriminator network. Next, the degree of distribution difference between the two sets of data is obtained through the comparative analysis of the discriminator. Then, a preliminary confidence value is calculated based on this degree of difference. Finally, the confidence value is dynamically corrected by combining the operating condition data of the turbocharger system to generate the final system coupling confidence parameters.

[0032] Continuing with the above case, once the system simultaneously obtains pressure spectrum energy data and vibration modal data, the conditional generative adversarial network processing in this step can establish a quantitative correlation model between pressure fluctuations and surface vibrations, and generate a confidence parameter reflecting the coupling strength between the two. This parameter provides an important decision-making basis for subsequent distributed optimization control.

[0033] S104. Based on the system coupling confidence parameters, the alternating direction multiplier method is used to perform distributed optimization of the operating points of multiple electronic water pumps and generate cooperative control commands. In this step, the alternating direction multiplier method refers to a mathematical computation method for solving distributed optimization problems; The operating point of a multi-electro-pump refers to the combination of operating parameters when multiple pumps are running in conjunction. Distributed optimization refers to the process of decomposing a global optimization problem into multiple subproblems and solving them in parallel. Coordinated control commands refer to a set of control parameters that coordinate the operation of multiple water pumps.

[0034] In this embodiment, the system coupling confidence parameters are first transformed into constraints of the optimization problem. Then, the multi-pump cooperative control problem is decomposed into several interrelated sub-optimization problems. Next, the sub-optimization problems are solved in parallel using the alternating direction multiplier method. Then, the solutions of each sub-problem are coordinated to form a global optimization solution. Finally, the optimization solution is transformed into specific executable control instructions.

[0035] Continuing with the above case, based on the system coupling confidence parameters obtained in the previous step, this step uses a distributed optimization algorithm to calculate the combination of operating parameters for each electronic water pump, generating control commands that enable efficient and coordinated operation, ensuring that the cooling system can respond to temperature change requirements in a timely manner.

[0036] S105. Based on the aforementioned collaborative control command, a PWM signal sequence is constructed using pulse width modulation technology to generate a control signal; In this step, pulse width modulation (PWM) technology refers to a technique for controlling power by adjusting the pulse width. A PWM signal sequence refers to a combination of pulse signals with adjustable width. Control signals refer to the specific electrical signals that drive the actuator of the electronic water pump.

[0037] In this embodiment, the target speed value of each water pump is first determined according to the cooperative control command, then the speed value is converted into the corresponding pulse width modulation parameter, then a pulse width modulation signal sequence is generated according to the time sequence, and finally the signal is amplified to form a control signal that can directly drive the water pump motor.

[0038] Continuing with the above example, the optimized collaborative control commands are converted into specific PWM drive signals. By precisely adjusting the pulse width, the operating speed of each water pump is controlled, thereby achieving fine adjustment of the coolant flow rate and meeting the cooling requirements of the system under different operating conditions.

[0039] S106. Based on the LIN / PWM communication bus, the control signal is transmitted using time-division multiple access to generate multi-mode coordinated control signals for the electronic water pump.

[0040] In this step, the LIN / PWM communication bus refers to a hybrid communication protocol that combines local interconnection networks and pulse width modulation. Time Division Multiple Access (TDMA) is a communication method that divides time into multiple time slots and allocates them to different devices. The multi-mode coordinated control signal for electronic water pumps refers to a comprehensive instruction set that enables coordinated control of multiple operating modes.

[0041] In this embodiment, the control signal is first encapsulated according to the format specified by the communication protocol. Then, a dedicated transmission time slot is allocated to each pump controller according to the time division multiple access protocol. Next, the encapsulated signal is distributed to each pump controller in sequence through the LIN bus. Then, the confirmation feedback information returned by each node is received. Finally, after confirming that all signals have been transmitted correctly, a complete cooperative control signal is generated.

[0042] Continuing with the above example, the generated control signals are reliably transmitted to each electronic water pump controller through a LIN / PWM hybrid communication method, ensuring that multiple water pumps can work together according to the optimized calculation results, achieving rapid response and precise control in scenarios of sudden temperature changes, thereby ensuring the stable operation of the turbocharger system.

[0043] In summary, S101 to S106, through multi-physics signal acquisition and fusion processing, combined with intelligent algorithms and distributed optimization control, ultimately achieve coordinated operation of the actuators via a dedicated communication protocol. This entire solution effectively addresses the control challenges in turbocharger systems under sudden temperature changes in the cooler, improves system response speed and control accuracy, enhances the reliability and adaptability of the vehicle's thermal management system, and provides an innovative solution for automotive engine thermal management.

[0044] To further improve the accuracy and response speed of cooler temperature control in turbocharger systems, in some embodiments, as described in S103, based on the spectral energy information and the vibration mode information, a conditional generative adversarial network is used to establish the correlation between vibration and pressure pulsation, generating system coupling confidence parameters, including: S201. Perform frequency band energy recombination processing on the spectral energy information to obtain recombined energy data; In S201, band energy recombination processing refers to the process of recombining and optimizing the band distribution in the spectrum energy information; Recombined energy data refers to energy distribution datasets with optimized characteristics formed after recombination.

[0045] In this embodiment, the energy values ​​of each frequency band contained in the spectral energy information are first standardized and preprocessed. Then, clustering is performed based on the energy correlation characteristics between frequency bands. Next, the energy values ​​of the frequency bands within the same group are weighted and summed. Then, the weighting coefficients of the energy distribution are adjusted according to the importance of different groups. Finally, reconstructed energy data with enhanced feature representation capabilities is generated. This process ensures the prominent performance of key frequency band energies, providing high-quality input for subsequent modal simulations.

[0046] S202. Perform mode shape node analysis on the vibration mode information to obtain analytical feature data; In S202, mode node analysis processing refers to the process of refining and deeply analyzing the vibration node features contained in the vibration mode information; Analytical feature data refers to the set of feature parameters with clear physical meaning extracted from vibration modes.

[0047] In this embodiment, the distribution of all key node locations in the vibration modal information is first identified. Then, the vibration amplitude and phase characteristic parameters of each node are calculated. Next, the vibration energy transfer relationship and coupling characteristics between the nodes are analyzed. Then, the natural frequency characteristics and damping characteristic parameters of the nodes are extracted. Finally, an analytical feature dataset containing complete node feature information is formed. This process ensures the comprehensiveness and accuracy of the vibration features.

[0048] S203. The recombined energy data is processed by vibration mode simulation using a conditional generation network to obtain simulated mode data; In S203, a conditional generation network refers to a deep neural network structure that can generate corresponding simulated data based on specific input conditional data; Vibration modal simulation processing refers to the process of generating highly realistic simulated vibration data by utilizing the powerful generative capabilities of generative networks; Simulated modal data refers to a dataset of simulated vibration features output by a generator network that is highly similar to real vibration data.

[0049] In this embodiment, the reconstructed energy data is first input as a conditional input vector into the input layer of the generator network. Then, the input features are extracted and abstracted layer by layer through multi-layer convolutional operations of the generator network. Next, deconvolutional layers are used for feature reconstruction and simulated data generation. Then, batch normalization layers are used to standardize the generated data. Finally, the output layer generates simulated modal data that is highly similar to the distribution of real vibration data. This process ensures the authenticity and usability of the simulated data.

[0050] S204. Input the simulated modal data and the analytical feature data into the discrimination network, perform distribution consistency discrimination processing, and obtain the distribution difference index; In S204, the discriminant network refers to a deep neural network structure specifically designed to distinguish the differences between real data and generated data; Distribution consistency discrimination refers to the process of accurately judging and analyzing the distribution similarity of two sets of data through a discriminant network; The distribution difference index is a comprehensive evaluation value that quantifies the degree of distribution difference between two sets of data.

[0051] In this embodiment, simulated modal data and analytical feature data are first simultaneously input into the input layer of the discriminant network. Then, the feature distance and similarity between the two sets of data are calculated using the multilayer perceptron structure of the discriminant network. Next, a specialized distance metric function is used to accurately calculate the degree of distribution difference. Then, an activation function maps the degree of difference to an interpretable score. Finally, an index value that accurately reflects the degree of distribution difference between the two sets of data is output. This process ensures the accuracy and reliability of the difference assessment.

[0052] S205. Perform confidence level transformation on the distribution difference index to obtain preliminary confidence parameters; In S205, confidence transformation processing refers to the mathematical process of converting distribution difference indicators into quantifiable confidence values. The initial confidence parameter refers to the initial confidence level value that has undergone preliminary transformation but has not yet been corrected.

[0053] In this embodiment, the distribution difference index is first preprocessed by standardization and normalization. Then, the difference values ​​are mapped to a probability range of zero to one using a sigmoid activation function. Next, the baseline confidence value is calculated based on the mapping result. Then, numerical smoothing and denoising are performed using a moving average window. Finally, stable and reliable preliminary confidence parameters are generated. This process ensures the rationality and stability of the confidence values.

[0054] S206. Based on the operating data of the vehicle turbocharger system, the preliminary confidence parameters are dynamically corrected to generate system coupling confidence parameters.

[0055] In S206, dynamic correction processing refers to the process of adaptively adjusting and optimizing parameters based on the collected operating condition data; The system coupling confidence parameter refers to the final confidence parameter obtained after correction of the operating condition data.

[0056] In this embodiment, operating condition data such as speed and load signals of the turbocharger system are first collected. Next, a dynamic correction relationship model between the operating condition data and confidence parameters is established. Then, corresponding correction coefficients and adjustment weights are calculated based on the current operating condition data. The preliminary confidence parameters are then weighted and optimized. Finally, system coupled confidence parameters that can adapt to specific operating condition changes are generated. This process ensures the environmental adaptability and reliability of the parameters.

[0057] Here is a specific example: When a vehicle undergoes rapid acceleration under high-temperature conditions, the turbocharger system enters a high-load operating state. The system first performs frequency band energy reconstruction processing on the acquired pressure signal spectrum to enhance the energy characteristics of key frequency bands. Simultaneously, it performs mode shape node analysis processing on the intercooler vibration mode information to extract important vibration characteristic parameters. Using a conditional generation network, high-quality simulated vibration mode data is generated based on the reconstructed energy data. Then, a discriminant network compares the distribution differences between the simulated data and the actual analytical characteristic data to obtain an accurate distribution difference index. This index is converted into a preliminary confidence parameter, which is then dynamically corrected and optimized based on current engine speed and intake air temperature data. Finally, a confidence parameter that accurately reflects the actual coupling state of the system is generated. The entire processing is completed in a very short time, ensuring the system can respond promptly to temperature changes.

[0058] In summary, S201 to S206, through a multi-step refined processing procedure, achieve accurate quantification and characterization of the multi-physics coupling relationship between pressure and vibration. By utilizing the collaborative working mechanism of condition generation network and discriminant network, the accuracy and reliability of coupling relationship identification are significantly improved. Combined with dynamic correction processing of operating condition data, the system's adaptability and robustness under different operating environment conditions are ensured, providing reliable parameter basis and support for subsequent distributed optimization control. This significantly improves the accuracy and response speed of cooler temperature control in the turbocharger system, enhancing the performance and reliability of the entire vehicle thermal management system.

[0059] To accurately acquire the vibration characteristics of the intercooler surface and improve temperature control accuracy, in some embodiments, as described in S102, a laser beam is emitted onto the vehicle intercooler surface using a helium-neon laser, and the vibration signal of the vehicle intercooler surface is inverted based on the changes in interference fringes to generate vibration mode information, including: S301. Using the helium-neon laser, laser beam projection processing is performed on the surface of the vehicle intercooler to obtain a reflected laser signal; In S301, laser beam projection processing refers to the optical processing process of generating a laser beam of a specific wavelength using a helium-neon laser and precisely guiding it to a designated area on the surface of the intercooler. Reflected laser signals are optical signals that carry surface vibration characteristics after being reflected from the surface of the intercooler.

[0060] In this embodiment, the helium-neon laser is first powered on to generate a stable laser beam. Then, an optical beam expander is used to enlarge the laser beam's diameter to obtain a suitable irradiation range. Next, a precision optical mirror is used to precisely guide the expanded laser beam to the center of the test area on the intercooler surface. A high-sensitivity photodetector is then used to receive the laser signal reflected from the surface. Finally, the received optical signal undergoes photoelectric conversion and preliminary amplification to obtain a reflected laser signal suitable for subsequent analysis. This process ensures accurate laser beam projection and high-quality acquisition of the reflected signal.

[0061] S302. Perform interference fringe generation processing on the reflected laser signal to obtain a pattern of alternating bright and dark fringes; In S302, interference fringe generation refers to the physical process of using the principle of optical interference to superimpose the reflected laser and the reference laser to generate an interference pattern; The alternating light and dark stripe pattern refers to an optical pattern with periodic changes in brightness formed by the interference of light waves.

[0062] In this embodiment, the reflected laser signal is first split into two optical signals by a beam splitter. Then, one of the optical signals is reflected by a reference optical path and re-combined with the other signal. Next, the phase difference between the combined optical paths is adjusted to satisfy the interference condition. Then, an imaging lens projects the interference field onto the target surface of a high-speed image sensor. Finally, an image acquisition system records a clear pattern of alternating bright and dark fringes. This process ensures the clarity and stability of the interference fringes.

[0063] S303. Continuously acquire and process the light and dark alternating stripe pattern to obtain a stripe dynamic sequence; In S303, continuous acquisition processing refers to the process of continuously capturing multiple frames of images of dynamically changing interference fringe patterns using a high-speed image acquisition system. A fringe dynamic sequence refers to a set of consecutive multi-frame interference fringe images arranged in chronological order.

[0064] In this embodiment, the acquisition parameters of the high-speed camera are first set, including frame rate and exposure time. Then, the image acquisition system is activated to continuously capture images of the interference fringe pattern. Next, the acquired image data is transmitted to the image processing unit for buffering. Then, timestamp information and sequence numbers are added to each frame. Finally, the data is organized chronologically to form a complete dynamic fringe sequence dataset. This process ensures the continuity and integrity of image acquisition.

[0065] S304. Based on the variation law of stripe spacing, the stripe dynamic sequence is subjected to spacing quantization processing to obtain stripe spacing data; In S304, spacing quantization refers to the process of accurately measuring and quantitatively analyzing the fringe spacing in an interference fringe image; Stripe spacing data refers to a numerical dataset that represents the size of the stripe spacing and its variation patterns.

[0066] In this embodiment, each frame of the fringe dynamic sequence is first preprocessed, including denoising and enhancement. Next, an edge detection algorithm is used to accurately identify the boundary position of each interference fringe. Then, the pixel distance between the center points of adjacent fringes is calculated and converted into actual physical distance. A mathematical model of the fringe spacing changing over time is then established. Finally, a quantized dataset containing the spacing values ​​and trends is generated. This process achieves accurate quantization of the fringe spacing.

[0067] S305. Based on the preset laser wavelength parameters, the stripe spacing data is processed by displacement conversion to obtain the surface micro-displacement. In S305, displacement conversion processing refers to the calculation process of converting fringe spacing data into actual vibration displacement values ​​based on the principle of optical interference. Surface micro-displacement refers to the precise quantification of the minute displacement generated on the surface of the intercooler during vibration.

[0068] In this embodiment, a conversion formula between fringe spacing and displacement value is first established based on the wavelength parameters of the helium-neon laser. Next, the fringe spacing data is standardized and normalized. Then, the displacement corresponding to each spacing value is calculated using interferometry formulas. Finally, the calculated displacement data is filtered and smoothed to eliminate noise, resulting in a precise dataset of surface micro-displacements. This process ensures the accuracy of the displacement calculation.

[0069] S306. The surface micro-displacement is processed by modal decomposition using a mode reconstruction algorithm to obtain vibration mode information.

[0070] In S306, the modal reconstruction algorithm refers to the mathematical calculation method that reconstructs discrete displacement data into continuous vibration modes; Modal decomposition processing refers to the signal processing process of decomposing a composite vibration signal into modal components of various orders; Vibration modal information refers to a complete dataset of vibration characteristics that includes parameters such as the frequency, damping ratio, and mode shape of each modal.

[0071] In this embodiment, the surface micro-displacement data is first transformed in the frequency domain to obtain spectral characteristics. Then, a modal parameter identification algorithm is used to identify the frequency and damping parameters of each mode. Next, a mode shape fitting algorithm is used to reconstruct the mode shape curves of each mode. The identified modal parameters are then verified and optimized. Finally, vibration modal information containing complete modal parameters is generated. This process achieves a complete conversion from displacement data to modal information.

[0072] Here is a specific example: When a vehicle accelerates rapidly in a high-temperature environment, the intercooler surface in the turbocharger system experiences complex vibrations due to thermal stress. The system first emits a laser beam from a helium-neon laser onto the intercooler surface, and the beam, after reflection, acquires the laser signal carrying vibration information. This signal interferes with a reference laser to form bright and dark fringes, which are continuously captured by a high-speed camera to obtain a dynamic fringe sequence. The image processing system analyzes the changes in fringe spacing and quantifies the data, converting the spacing data into precise micro-displacements based on the laser wavelength parameters. Finally, a mode shape reconstruction algorithm decomposes the displacement data into modal parameters of each order, obtaining complete vibration modal information, providing accurate input for subsequent multiphysics coupling analysis.

[0073] In summary, S301 to S306, through a complete optical measurement and signal processing chain, achieve non-contact, high-precision measurement of intercooler surface vibration. They utilize laser interferometry to capture microscopic vibration characteristics and extract accurate modal parameters through advanced signal processing algorithms, providing reliable vibration data support for multi-physics coupling analysis. This significantly improves the accuracy and reliability of vibration measurement and lays a solid technical foundation for intelligent thermal management of turbocharger systems.

[0074] To optimize the coordinated control performance of multiple electric water pumps in a turbocharger system, in some embodiments, as described in S104, based on the system coupling confidence parameters, a distributed optimization of the operating points of the multiple electric water pumps is performed using the alternating direction multiplier method to generate coordinated control commands, including: S401. Decompose the multi-pump control constraint conditions of the system coupling confidence parameters to obtain a distributed constraint set. In S401, the decomposition of multi-pump control constraints refers to the process of decomposing the system-level global optimization problem into multiple sub-problems according to the physical characteristics and operating limitations of each pump. A distributed constraint set is a complete set that contains independent constraints for each subproblem.

[0075] In this embodiment, the multi-pump collaborative working constraint relationship implied in the system coupling confidence parameters is first analyzed in depth, including global constraints such as flow balance constraints, pressure balance constraints, and power limitation constraints. Secondly, based on the rated operating parameters, operating states, and performance characteristics of each pump, the complex global constraints are decomposed into simplified constraints applicable to individual pumps. Next, corresponding mathematical expressions for the constraints are established for each independent pump, clarifying its upper and lower limits of operating parameters and performance boundaries. Then, the coordination and feasibility between the various sub-constraints are verified using a constraint compatibility analysis algorithm, ensuring that the decomposed constraint set maintains the integrity and consistency of the global constraints. Finally, a distributed constraint set containing the independent constraints of all pumps is generated, laying the foundation for subsequent distributed optimization calculations.

[0076] S402. Set an initial working point for the distributed constraint set to obtain an initial working state set; In S402, the initial operating point setting process refers to the process of providing appropriate initial parameter configurations for the optimization calculations of each pump. The initial operating state set refers to a complete dataset containing the initial operating parameter configurations of all water pumps.

[0077] In this embodiment, historical operating data of each water pump is first collected, including performance parameters and optimization records under typical operating conditions. Next, the current operating status of the turbocharger system is considered, including cooling demand intensity, system pressure level, and temperature distribution. Then, an experience-based heuristic method is used to set reasonable initial operating parameters for each water pump, including initial speed setpoints, expected flow rates, and power allocation ratios. Finally, a constraint satisfaction verification algorithm is used to verify whether the initial parameters satisfy all constraints in the distributed constraint set. For initial parameter configurations that do not meet the constraints, a constraint relaxation-based adjustment method is used for parameter optimization and correction. Finally, a set of feasible and near-optimal initial operating states is generated.

[0078] S403. Perform parallel optimization calculations on the initial working state set to obtain a local optimization result set; In S403, parallel optimization computation refers to the computation process in which each pump controller performs independent optimization solutions simultaneously based on local constraints. The local optimization result set refers to the data set of optimization results calculated by the local optimizer of each water pump.

[0079] In this embodiment, the corresponding constraints in the distributed constraint set are first distributed to the local controllers of each water pump, ensuring that each controller obtains its own exclusive optimization constraints. Next, parallel optimization computation processes are initiated in each local controller, using the same optimization algorithm but different initial parameters for independent solution. Then, the parallel computing capabilities of multi-core processors are utilized to execute multiple optimization computation tasks simultaneously, with each task responsible for the local optimization problem of one water pump. During the optimization computation, the optimization process status of each local controller is monitored to ensure the stability and convergence of the computation process. Finally, the optimization computation results of all local controllers are collected, including optimal operating parameters, objective function values, and constraint satisfaction status, forming a complete local optimization result set.

[0080] S404. Through the communication network of the LIN / PWM communication bus, the local optimization result set is processed by intermediate variable exchange to obtain global coordination variables; In S404, intermediate variable exchange processing refers to the process of exchanging intermediate coordination variables generated during the optimization calculation through the communication network among the various pump controllers. Global coordination variables refer to a set of shared parameters used to coordinate the optimization direction of various water pumps and maintain global consistency.

[0081] In this embodiment, a data exchange protocol based on the LIN / PWM communication bus is first established, defining the data format and transmission specifications of intermediate variables. Next, each pump controller encapsulates and packages the intermediate result variables generated by local optimization calculations according to the protocol format. Then, using a time-division multiple access (TDMA) communication scheduling method, data packets are sent to the shared communication network within a specified communication time slot. Each controller then receives intermediate variable data packets sent by other controllers, performs data parsing and integrity verification. Finally, based on all received intermediate variable information, a weighted average algorithm or a consensus algorithm is used to calculate and generate a global coordination variable.

[0082] S405. Perform working point iterative update processing on the global coordination variables to obtain the updated working state set; In S405, the operating point iterative update process refers to the process of iteratively optimizing and improving the operating parameters of each pump based on global coordination variables. The updated working state set refers to the improved set of working parameters obtained after iterative optimization.

[0083] In this embodiment, firstly, based on the system coordination information contained in the global coordination variables, the adjustment direction and step size of each pump's operating parameters are calculated; secondly, an iterative optimization algorithm based on gradient descent is used to perform the first iteration update of the initial operating state, generating a new operating parameter configuration; then, the constraint satisfaction verification is performed to check whether the updated operating parameters meet all constraint requirements; for operating parameter configurations that violate constraints, the projection gradient method or constraint penalty function method is used to correct and adjust the parameters; then, the degree of improvement of the objective function of the updated operating state is calculated to evaluate the effect of iterative optimization; finally, an updated set of operating states is generated to prepare for the next iteration optimization or the final result output.

[0084] S406. Based on the preset convergence judgment conditions, the updated working state set is judged and processed to generate a collaborative control command.

[0085] In S406, the convergence criterion refers to the criteria used to determine whether the distributed optimization process has reached a convergence state and can terminate the iteration. Judgment processing refers to the analytical process of performing convergence tests and evaluations on the current optimization results; Coordinated control commands refer to the final set of control commands generated to coordinate the coordinated operation of multiple water pumps.

[0086] In this embodiment, firstly, convergence criteria parameters are set, including the objective function change threshold, variable change tolerance range, and maximum number of iterations. Secondly, the degree of difference between the working state set of the current iteration and the working state set of the previous iteration is calculated, including the parameter change and objective function improvement. Next, the current degree of difference is compared and analyzed with the preset convergence tolerance range to determine whether the convergence conditions are met. For optimization results that meet the convergence conditions, the final control instruction set is generated, including the optimal operating parameters and control commands for each pump. For results that do not meet the convergence conditions, a new round of iterative optimization is started, and the optimization algorithm parameters are dynamically adjusted according to the current convergence state. Finally, collaborative control instructions are output to ensure that multiple pumps can work collaboratively according to the optimization results.

[0087] Here is a specific example: When the turbocharger system detects a sudden change in intercooler temperature, the control system first decomposes the complex multi-pump coordinated control problem into multiple sub-problems based on the acquired system coupling confidence parameters, and sets an initial operating point for each pump based on its current operating state. Each pump controller performs parallel local optimization calculations, exchanging intermediate optimization results via a LIN / PWM communication network to generate global coordination variables. Based on the coordination variable information, the system iteratively updates the operating points of each pump multiple times, verifying the convergence conditions after each iteration until the accuracy requirements are met. Finally, it generates the optimal control command to coordinate the operation of multiple pumps, achieving rapid response and precise control under sudden temperature changes, ensuring the stable and efficient operation of the turbocharger system.

[0088] In summary, S401 to S406, through a systematic distributed optimization framework, achieve collaborative optimization calculation of multiple electric pump operating points, employ constraint decomposition to handle complex system constraints, utilize parallel computing to improve optimization efficiency, ensure global coordination through intermediate variable exchange, and approximate the optimal solution through iterative optimization, ultimately generating high-quality collaborative control commands. This significantly improves the control accuracy and response speed of the multi-pump system, provides an intelligent thermal management solution for turbocharged systems, and enhances the system's reliability and adaptability.

[0089] To improve the accuracy and reliability of vibration modal simulation, in some embodiments, as described in S203, a conditional generation network is used to perform vibration modal simulation processing on the recombined energy data to obtain simulated modal data, including: S501. Perform a pressure feature encoding operation on the recombined energy data to obtain a pressure feature vector; In S501, the pressure feature encoding operation refers to the data processing process of converting the pressure feature information in the recombined energy data into a vector representation; A pressure feature vector is a numerical vector representation that contains quantitative information about pressure features.

[0090] In this embodiment, the multi-dimensional pressure features in the recombined energy data are first standardized and preprocessed to eliminate the influence of different dimensions. Then, deep feature information in the pressure signal is extracted layer by layer through a multi-layer convolutional neural network structure. Next, the extracted features are reduced in dimensionality and key feature selection is performed by a max pooling layer. Then, the filtered features are mapped to a fixed-dimensional vector representation space through a fully connected neural network layer. Finally, a feature vector representation containing rich pressure feature information is generated.

[0091] S502. Perform a latent space transformation operation on the pressure feature vector to obtain a latent feature representation; In S502, the latent space transformation operation refers to the mathematical transformation process that maps the pressure feature vector to the latent feature space. Latent feature representation refers to a low-dimensional representation in the latent space that can better express the essential features of data.

[0092] In this embodiment, a deep encoder network structure is first constructed and a suitable number of network layers and neurons are designed. Then, the pressure feature vector is input into the input layer of the encoder network through the forward propagation algorithm. Next, a nonlinear activation function is used to introduce complex feature transformation relationships in the network. Then, the high-dimensional features are mapped to the low-dimensional latent space through the step-by-step transformation of the multi-layer neural network. Finally, a latent feature representation that can capture the essential features and inherent laws of the data is obtained.

[0093] S503. Combine the operating data of the vehicle turbocharger system to perform a conditional constraint operation on the potential feature representation to obtain a conditional feature representation; In S503, condition constraint operation refers to the process of introducing working condition data as condition information into the feature representation; Conditional feature representation refers to an enhanced feature representation that incorporates operating condition information.

[0094] In this embodiment, various operating condition data of the turbocharger system are first collected, including key operating parameters such as speed, temperature, and pressure. Then, the collected operating condition data are standardized and feature encoded. Next, the processed operating condition information and potential feature representation are deeply fused through a feature fusion layer. Then, an attention mechanism is used to dynamically adjust the feature weight allocation under different operating conditions. Finally, a condition feature representation containing rich condition information and highly adapted to the operating conditions is generated.

[0095] S504. Perform a nonlinear transformation operation on the conditional feature representation through a multi-layer perceptron architecture to obtain transformed feature data; In S504, nonlinear transformation operation refers to the process of implementing complex nonlinear transformations through multi-layer neural networks; Transformed feature data refers to enhanced feature data obtained after nonlinear transformation.

[0096] In this embodiment, a deep multilayer perceptron network structure is first constructed, which includes multiple hidden layers and corresponding output layers. Then, the conditional feature representation is input into the network input layer through the forward propagation algorithm. Next, various activation functions are used to introduce the necessary nonlinear transformation capability into the network. Then, deeper and more abstract feature representations are extracted through the layer-by-layer nonlinear transformation of the deep neural network. Finally, transformed feature data with stronger expressive power and discriminative power are obtained.

[0097] S505. Perform a spatial reconstruction operation on the transformed feature data to obtain preliminary modal data; In S505, spatial reconstruction operation refers to the process of reorganizing feature data into a spatial distribution structure; Preliminary modal data refers to the preliminary results of modal data with spatial structural characteristics.

[0098] In this embodiment, the spatial dimension and organizational structure requirements of the target modal data are first determined. Then, the transformed feature data is upsampled and spatially expanded through a deconvolutional neural network layer. Next, the spatial dimension reconstruction and structural adjustment of the feature map are achieved using transposed convolution. Then, feature information at different scales and levels is fused through a skip connection mechanism. Finally, preliminary modal data with complete spatial structure features is generated.

[0099] S506. Perform detail enhancement operation on the preliminary modal data to obtain simulated modal data.

[0100] In S506, detail enhancement refers to the process of refining and enhancing the features of the preliminary modal data; simulated modal data refers to the final generated simulated vibration modal data with high fidelity.

[0101] In this embodiment, a comprehensive feature analysis and quality assessment are first performed on the preliminary modal data. Then, a deep residual learning network is used to enhance the performance of detailed features and the clarity of contours. Next, a generative adversarial network is used to refine the data and improve its quality. Then, a feature optimization algorithm is used to further improve the accuracy and consistency of the data. Finally, simulated modal data with high-quality detailed features and a distribution highly consistent with the real modal data is generated.

[0102] Here is a specific example: When a vehicle undergoes rapid acceleration under high temperature and high load conditions, the turbocharger system generates specific pressure fluctuation characteristics. The system first performs deep pressure feature encoding on the reconstructed energy data, extracting key feature information and converting it into a feature vector representation. Subsequently, a more abstract and essential feature representation is obtained through latent space transformation, and combined with the collected operating condition data to generate a feature representation with conditional constraints. After a complex nonlinear transformation using a multilayer perceptron, the feature data is significantly enhanced and reconstructed into preliminary modal data with spatial structure. Finally, through detail enhancement processing, high-precision, high-fidelity simulated modal data is generated, providing accurate and reliable data input for subsequent vibration characteristic analysis and multiphysics coupling.

[0103] In summary, S501 to S506 achieve accurate simulation and conversion from pressure energy data to vibration modal data through a multi-step, refined processing flow. They effectively capture complex nonlinear mapping relationships using deep learning technology, enhance the accuracy and adaptability of the simulation by combining operating condition data, and ensure data quality and reliability through spatial reconstruction and detail enhancement. This provides high-quality simulation data support for multiphysics coupling analysis, significantly improves the accuracy and practicality of vibration modal simulation, and provides an important technical foundation for the intelligent control of turbocharger systems.

[0104] To accurately acquire vibration information of the intercooler surface and improve temperature control accuracy, in some embodiments, as described in S301, the helium-neon laser is used to project a laser beam onto the surface of the vehicle intercooler to obtain a reflected laser signal, including: S601. A single-wavelength laser beam is generated using the helium-neon laser. In S601, a helium-neon laser refers to an optical emitting device capable of generating laser light of a specific wavelength. A single-wavelength laser beam is a laser beam that has a single wavelength characteristic.

[0105] In this embodiment, the power supply system of the helium-neon laser is first started and waited for it to reach a thermally stable state. Then, the discharge current of the laser tube is adjusted by a precision current control device to make it work at the operating point. Next, the mode selection characteristics of the laser resonator are used to filter out pure single transverse mode and single longitudinal mode laser outputs. Then, the laser energy in the cavity is exported through the output coupling mirror to form a stable single-wavelength laser beam. Finally, a laser power meter is used to monitor the stability of the output power and make corresponding adjustments.

[0106] S602. The single-wavelength laser beam is shaped by an optical lens group to form a uniform light spot; In S602, an optical lens group refers to an optical system composed of multiple optical lenses arranged in a specific order; Beam shaping refers to the process of optimizing the spatial distribution and wavefront characteristics of a laser beam. A uniform light spot refers to an optical spot with a uniform light intensity distribution and a flat wavefront.

[0107] In this embodiment, a single-wavelength laser beam is first introduced into a beam expansion system composed of convex and concave lenses for collimation and beam expansion. Then, a spatial filter is used to filter out higher-order modes and stray light components in the laser beam. Next, a beam shaper is used to convert the Gaussian intensity distribution into a flat-top distribution. Then, a focusing lens group is used to focus the shaped beam to the required size. Finally, a regular light spot with uniform light intensity distribution and minimal wavefront distortion is formed on the target plane.

[0108] S603. A beam splitter is used to divide the uniform light spot into a reference beam and a measurement beam; In S603, a beam splitter refers to an optical element that achieves beam splitting based on the principle of optical thin-film interference. The reference beam refers to the reference beam used for interferometric reference; The measuring beam refers to the working beam used for actual measurement.

[0109] In this embodiment, the uniform light spot is first precisely aligned with the incident surface of the beam splitter and the incident angle is ensured to meet the design requirements. Then, the incident beam is split into two beams with a specific splitting ratio through the special coating surface of the beam splitter. Next, the azimuth angle of the beam splitter is adjusted so that the two outgoing beams maintain a specific spatial separation angle. Then, the optical paths of the two beams are independently adjusted to ensure the parallelism of the optical paths and the quality of the light spot. Finally, a reference beam and a measurement beam with a stable power ratio and good optical quality are obtained.

[0110] S604. The measuring beam is directed to a specific measuring point on the surface of the vehicle intercooler using a set of reflectors. In S604, the mirror assembly refers to an optical path guiding system composed of multiple highly reflective mirrors arranged in a precise manner; A specific measurement point refers to the precisely determined center position of the area to be measured on the surface of the intercooler.

[0111] In this embodiment, the measurement point position is first determined based on the geometric features and vibration characteristics of the intercooler surface. Then, the reflector group is fixed by a six-dimensional precision adjustment frame and the spatial attitude of each reflector is precisely adjusted. Next, the measurement beam is guided into the first high-reflectivity mirror of the reflector group. Then, the beam is precisely projected onto the target measurement point through the continuous reflection and guidance of the reflector group. Finally, a high-precision spot analyzer is used to verify the spot position accuracy and perform nanometer-level fine-tuning correction.

[0112] S605. Receive the scattered light beam returned from the specific measurement point to obtain the original optical signal; In S605, the scattered beam refers to the optical signal carrying surface information that is diffusely reflected back from the surface of the measurement point. The raw optical signal refers to the initial optical signal without any processing.

[0113] In this embodiment, a large-aperture optical collection device is first set in the receiving optical path of the measurement beam. Then, the spatial position and receiving angle of the collection device are adjusted to maximize the collection efficiency. Next, a compound lens group is used to efficiently focus the scattered beam onto the sensitive area of ​​the photodetector. Then, a narrow-band optical filter is used to filter out ambient stray light interference. Finally, the received optical signal is converted into a high-quality electrical signal by a low-noise amplifier to obtain the original optical signal.

[0114] S606. The original optical signal is subjected to interference processing with the reference beam to obtain the reflected laser signal.

[0115] In S606, interference processing refers to the process of superimposing two coherent beams to generate an interference pattern and extracting phase information; Reflected laser signals refer to interferometric optical signals that contain complete vibration information.

[0116] In this embodiment, the optical path corresponding to the original optical signal is first precisely aligned with the optical path of the reference beam in space. Then, the optical path difference between the two beams is brought close to zero using a precision optical path adjustment device. Next, a polarizing beam splitter is used to re-merge the two beams to generate a stable interference pattern. Then, the interference optical signal is introduced into a high-sensitivity photodetector through the interferometer optical system. Finally, the electrical signal output by the detector is amplified with low noise and bandpass filtered to obtain a high-quality reflected laser signal.

[0117] Here is a specific example: When a vehicle accelerates rapidly under high temperature and high load conditions, the surface of the turbocharger intercooler experiences complex vibrations due to thermal stress. The system first activates a helium-neon laser to generate a stable single-wavelength laser beam, which is then shaped into a uniformly distributed energy spot by an optical lens group. A beam splitter divides the beam into a reference beam and a measurement beam, with the measurement beam precisely guided to a specific measurement point on the intercooler surface by a set of reflectors. After receiving the scattered light returning from the surface, interference processing is performed with the reference beam to ultimately obtain a reflected laser signal containing rich vibration information, providing a complete and accurate data foundation for subsequent vibration modal analysis.

[0118] In summary, S601 to 606 achieve non-contact, high-precision measurement of intercooler surface vibration through a complete and precise optical processing chain. The excellent monochromaticity and stability of the helium-neon laser ensure light source quality, a precision optical system enables accurate beam control and positioning, and interferometry effectively extracts microscopic vibration information. This provides reliable vibration data support for intelligent temperature control of the turbocharger system, significantly improving the accuracy and reliability of vibration measurement and providing crucial technical support for optimizing the vehicle's thermal management performance.

[0119] To improve the communication efficiency and system coordination of multi-electronic water pump collaborative control, in some embodiments, as described in S404, the local optimization result set is processed through the communication network of the LIN / PWM communication bus to obtain global coordination variables, including: S701. Using the communication network of the LIN / PWM communication bus, the local optimization result set is encapsulated to obtain a standardized data packet; In S701, data encapsulation processing refers to the process of converting and packaging the local optimization result set according to the requirements of the communication protocol. A standardized data packet is a data unit that conforms to communication protocol specifications and contains complete transmission information.

[0120] In this embodiment, the data structure and content composition of the local optimization result set are first analyzed to identify the key parameters and status information that need to be transmitted. Secondly, based on the frame structure design requirements of the LIN / PWM communication protocol, the composition format of the data packet's frame header, data field, and check field is determined. Next, corresponding identifiers and length information are added to each data item to ensure correct data parsing. Then, byte order conversion and encoding processing are performed on the data to ensure data compatibility between different nodes. Finally, error control information such as cyclic redundancy check codes is added to generate standardized data packets that fully comply with the communication protocol requirements.

[0121] S702. Based on the time division multiple access communication protocol, the standardized data packets are processed by time slot allocation to obtain a scheduled transmission sequence; In S702, the time division multiple access communication protocol refers to a multiple access control protocol that divides time resources into multiple time slots and allocates them to different nodes. Time slot allocation processing refers to the process of allocating transmission time resources to each node based on network status and service requirements; The scheduled transmission sequence refers to a scheduling plan that includes a time slot allocation scheme and transmission timing.

[0122] In this embodiment, information such as the number, priority, and traffic demand of all active nodes in the network is first obtained; then, based on the node status and data urgency, a dynamic time slot allocation algorithm is used to calculate the optimal time slot allocation scheme; next, a specific transmission time window is allocated to each node, including start time and duration parameters; then, considering transmission delay and jitter requirements, the guard interval setting between time slots is optimized; finally, a scheduled transmission sequence containing complete time slot allocation information and transmission timing requirements is generated.

[0123] S703. Through the physical link of the communication network, the scheduled transmission sequence is processed by data packet transmission to obtain the acknowledgment signal returned by each distributed node; In S703, a physical link refers to the actual signal transmission channel in a communication network; Data packet transmission processing refers to the complete process of sending and receiving data packets through a physical medium; An acknowledgment signal is a response message returned by the receiver after the data has been correctly received.

[0124] In this embodiment, the transmission time slot and parameter configuration of the current node are first determined according to the scheduled transmission sequence; then, the physical layer transmission circuit is activated within the specified time slot to convert the standardized data packets into electrical signals for modulation and driving; next, the data packets are transmitted through the physical link, while the link quality and signal strength are monitored; then, a receive timer is started to wait for the receiver's acknowledgment response; finally, the acknowledgment signal is received and identified within a fixed time, and the transmission status information is recorded.

[0125] S704. Receive the confirmation signals returned by each distributed node, perform transmission integrity verification processing, and obtain the data packets of each node; In S704, transmission integrity verification processing refers to the process of verifying the integrity and correctness of data transmission. Each node's data packet refers to the complete set of data packets successfully received from all nodes.

[0126] In this embodiment, the system first receives acknowledgment signals returned by each node and parses the status information and sequence number therein; secondly, it verifies the checksum field of each data packet to verify whether any errors occurred during data transmission; then, it compares the sequence numbers of the sent and received data packets to ensure the integrity and order of the data packets; next, it calculates performance indicators such as the transmission success rate and retransmission count of each node; finally, it organizes all successfully received data packets according to the node identifier to form a complete data packet set.

[0127] S705. Perform protocol parsing processing on the data packets of each node and extract intermediate variable data; In S705, protocol parsing refers to the process of decoding data packets and extracting their content according to the communication protocol specifications. Intermediate variable data refers to the optimization result data extracted from the data package for collaborative computing.

[0128] In this embodiment, firstly, the frame header information of each data packet is parsed according to the format specified by the communication protocol to obtain the length and type identifier of the data packet; secondly, the payload content in the data field is extracted according to the protocol specification, and byte order conversion and decoding are performed; then, the integrity and consistency of the data are verified, and the existence of transmission errors is verified; then, the parsed data is classified and organized according to the data identifier and restored to the original optimization result format; finally, an intermediate variable dataset containing the optimization results of all nodes is generated.

[0129] S706. Through a variable integration algorithm, the intermediate variable data is processed collaboratively to generate a global coordination variable.

[0130] In S706, the variable integration algorithm refers to a mathematical calculation method that coordinates the processing of multiple intermediate variables and generates a global coordinated value. Collaborative computing refers to a processing method that coordinates and optimizes intermediate results from multiple distributed nodes. Global coordination variables refer to the final optimization results generated after collaborative computation for global coordination.

[0131] In this embodiment, the correlation and coupling relationships between various intermediate variables are first analyzed to establish a mathematical model between the variables; second, a variable integration strategy based on a consensus algorithm is designed to ensure the fairness and optimality of global coordination; then, an iterative optimization method is used to calculate the coordination weight and adjustment amount of each variable; then, the correctness and feasibility of the calculation results are verified through a distributed consensus mechanism; finally, a global coordination variable that can coordinate the work of all nodes is generated.

[0132] Here is a specific example: When a vehicle is under rapid acceleration, the temperature of the turbocharger system changes drastically. Each electronic water pump controller first encapsulates its local optimization results into standard data packets according to the LIN / PWM communication protocol, and then transmits the data via the physical link within the allocated time slot according to the time-division multiple access protocol. After receiving all data packets, the main controller performs integrity verification to ensure error-free data transmission. It then parses the intermediate variable data and finally calculates and generates a global coordination variable using a variable integration algorithm. This variable guides each water pump to adjust its operating state, achieving rapid response and precise control of the cooling system, and ensuring stable operation of the turbocharger system under sudden temperature changes.

[0133] In summary, S701 to S706 achieve efficient data exchange and collaborative computing in the distributed optimization process of multiple water pumps through a complete communication processing flow. They utilize standard communication protocols and time-division multiple access mechanisms to ensure the orderliness and reliability of data transmission, guarantee data quality through integrity verification, and employ advanced variable integration algorithms to achieve global optimization coordination. This significantly improves the collaborative control performance and dynamic response capability of the multi-pump system, providing strong technical support for vehicle thermal management systems.

[0134] Figure 4 This application provides a schematic diagram of a specific implementation of a multi-mode collaborative control system for an electronic water pump based on LIN / PWM communication, as shown in the following embodiment. Figure 4 The system may include: Pressure spectrum module 41 is used to perform frequency band energy analysis on the pressure pulsation based on the pipeline pressure signal of the vehicle turbocharger system and generate spectrum energy information by using fast Fourier transform. The laser vibration measurement module 42 is used to emit a laser beam to the surface of the vehicle intercooler using a helium-neon laser, and invert the vibration signal of the vehicle intercooler surface based on the change of interference fringes to generate vibration mode information. The coupling confidence module 43 is used to establish the correlation between vibration and pressure pulsation based on the spectral energy information and the vibration mode information, combined with a conditional generative adversarial network, and generate system coupling confidence parameters. The collaborative optimization module 44 is used to perform distributed optimization of the operating points of multiple electronic water pumps based on the system coupling confidence parameters and using the alternating direction multiplier method to generate collaborative control commands. Signal generation module 45 is used to construct a PWM signal sequence based on the cooperative control command using pulse width modulation technology to generate a control signal; The bus transmission module 46 is used to transmit the control signal based on the LIN / PWM communication bus and in a time-division multiple access manner, and to generate multi-mode coordinated control signals for the electronic water pump.

[0135] This application provides an embodiment of a multi-mode cooperative control system for an electronic water pump based on LIN / PWM communication to implement the aforementioned multi-mode cooperative control method for an electronic water pump based on LIN / PWM communication. Therefore, the specific implementation of the multi-mode cooperative control system for an electronic water pump based on LIN / PWM communication can be found in the embodiment section of the preceding text on the multi-mode cooperative control method for an electronic water pump based on LIN / PWM communication. The specific implementation can be referred to the description of the corresponding embodiments, and will not be repeated here.

[0136] This application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for executing the computer program to implement the steps of the above-described LIN / PWM communication-based multi-mode cooperative control method for an electronic water pump.

[0137] This application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described multi-mode cooperative control method for an electronic water pump based on LIN / PWM communication.

[0138] In one exemplary embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory, random access memory, portable hard drives, magnetic disks, or optical disks.

[0139] Embodiments of the present invention also provide a computer program product, which includes a computer program that, when executed by a processor, implements the steps in any of the embodiments of the electronic water pump multi-mode cooperative control method based on LIN / PWM communication.

[0140] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0141] The foregoing has provided a detailed description of the multi-mode cooperative control method, system, electronic device, and storage medium for an electronic water pump based on LIN / PWM communication provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are merely for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A multi-mode cooperative control method for an electronic water pump based on LIN / PWM communication, characterized in that, include: Based on the pipeline pressure signal of the vehicle turbocharger system, the pressure pulsation is analyzed by frequency band energy using fast Fourier transform to generate spectral energy information. A laser beam is emitted onto the surface of the vehicle intercooler using a helium-neon laser, and vibration mode information is generated by inverting the vibration signal of the vehicle intercooler surface based on the changes in interference fringes. Based on the spectral energy information and the vibration mode information, a conditional generative adversarial network is used to establish the correlation between vibration and pressure pulsation, and system coupling confidence parameters are generated. Based on the system coupling confidence parameters, the alternating direction multiplier method is used to perform distributed optimization of the operating points of multiple electronic water pumps and generate cooperative control commands. Based on the aforementioned collaborative control instructions, a PWM signal sequence is constructed using pulse width modulation (PWM) technology to generate a control signal. Based on the LIN / PWM communication bus, the control signals are transmitted using time-division multiple access to generate multi-mode coordinated control signals for the electronic water pump.

2. The method according to claim 1, characterized in that, Based on the spectral energy information and the vibration mode information, a conditional generative adversarial network is used to establish the correlation between vibration and pressure pulsation, generating system coupling confidence parameters, including: The spectral energy information is subjected to band energy recombination processing to obtain recombined energy data; The vibration mode information is subjected to mode shape node analysis to obtain analytical feature data; The recombined energy data is processed using a conditional generation network to perform vibration mode simulation, resulting in simulated mode data. The simulated modal data and the analytical feature data are input into a discriminant network for distribution consistency discrimination processing to obtain a distribution difference index. The distribution difference index is subjected to confidence level transformation to obtain preliminary confidence parameters; Based on the operating data of the vehicle turbocharger system, the preliminary confidence parameters are dynamically corrected to generate system coupling confidence parameters.

3. The method according to claim 1, characterized in that, A laser beam is emitted onto the surface of the vehicle's intercooler using a helium-neon laser, and vibration mode information is generated by inverting the vibration signal of the intercooler surface based on changes in interference fringes, including: Using the helium-neon laser, a laser beam is projected onto the surface of the vehicle's intercooler to obtain a reflected laser signal; The reflected laser signal is subjected to interference fringe generation processing to obtain a pattern of alternating bright and dark fringes; The light and dark alternating stripe pattern is continuously acquired and processed to obtain a stripe dynamic sequence; Based on the variation law of stripe spacing, the stripe dynamic sequence is subjected to spacing quantization processing to obtain stripe spacing data; Based on the preset laser wavelength parameters, the stripe spacing data is processed by displacement conversion to obtain the surface micro-displacement. The surface micro-displacement is modally decomposed using a mode reconstruction algorithm to obtain vibration mode information.

4. The method according to claim 1, characterized in that, Based on the system coupling confidence parameters, the alternating direction multiplier method is used for distributed optimization of the operating points of multiple electronic water pumps to generate cooperative control commands, including: The system coupling confidence parameters are decomposed using multi-pump control constraints to obtain a distributed constraint set; An initial working point is set for the distributed constraint set to obtain an initial working state set; Parallel optimization computation is performed on the initial working state set to obtain a local optimization result set; The local optimization result set is processed by exchanging intermediate variables through the communication network of the LIN / PWM communication bus to obtain global coordination variables; The global coordination variables are updated iteratively at work points to obtain an updated work state set; Based on preset convergence criteria, the updated working state set is judged and processed to generate collaborative control instructions.

5. The method according to claim 2, characterized in that, The recombined energy data is processed using a conditional generation network to perform vibration mode simulation, resulting in simulated mode data, including: Perform a pressure feature encoding operation on the recombined energy data to obtain a pressure feature vector; Perform a latent space transformation operation on the pressure feature vector to obtain a latent feature representation; By combining the operating data of the vehicle turbocharger system, a conditional constraint operation is performed on the potential feature representation to obtain a conditional feature representation; A nonlinear transformation operation is performed on the conditional feature representation using a multi-layer perceptron architecture to obtain transformed feature data; Perform a spatial reconstruction operation on the transformed feature data to obtain preliminary modal data; Perform detail enhancement operations on the preliminary modal data to obtain simulated modal data.

6. The method according to claim 3, characterized in that, Using the helium-neon laser, a laser beam is projected onto the surface of the vehicle's intercooler to obtain a reflected laser signal, including: A single-wavelength laser beam is generated using the helium-neon laser described above. The single-wavelength laser beam is shaped using an optical lens group to form a uniform light spot; The uniform light spot is divided into a reference beam and a measurement beam using a beam splitter prism; The measurement beam is directed to a specific measurement point on the surface of the vehicle's intercooler using a set of mirrors. Receive the scattered light beam returned from the specific measurement point to obtain the original optical signal; The original optical signal is subjected to interference processing with the reference beam to obtain the reflected laser signal.

7. The method according to claim 4, characterized in that, Through the communication network of the LIN / PWM communication bus, intermediate variable exchange processing is performed on the local optimization result set to obtain global coordination variables, including: Using the communication network of the LIN / PWM communication bus, the local optimization result set is encapsulated to obtain a standardized data packet; Based on the time division multiple access communication protocol, the standardized data packets are processed by time slot allocation to obtain a scheduled transmission sequence; Through the physical link of the communication network, the scheduled transmission sequence is processed into data packets to obtain acknowledgment signals returned by each distributed node; The system receives confirmation signals from each distributed node, performs transmission integrity verification, and obtains data packets from each node. The data packets of each node are processed by protocol parsing to extract intermediate variable data; The intermediate variable data is processed collaboratively using a variable integration algorithm to generate a globally coordinated variable.

8. A multi-mode cooperative control system for an electronic water pump based on LIN / PWM communication, characterized in that, include: The pressure spectrum module is used to perform frequency band energy analysis on the pipeline pressure signal based on the vehicle turbocharger system, and generate spectrum energy information by using fast Fourier transform to analyze the pressure pulsation. The laser vibration measurement module is used to emit a laser beam to the surface of the vehicle intercooler using a helium-neon laser, and invert the vibration signal of the vehicle intercooler surface based on the changes in interference fringes to generate vibration mode information; The coupling confidence module is used to establish the correlation between vibration and pressure pulsation based on the spectral energy information and the vibration mode information, combined with a conditional generative adversarial network, and to generate system coupling confidence parameters. The collaborative optimization module is used to perform distributed optimization of the operating points of multiple electronic water pumps based on the system coupling confidence parameters and using the alternating direction multiplier method to generate collaborative control commands. The signal generation module is used to construct a PWM signal sequence based on the cooperative control command using pulse width modulation technology to generate a control signal; The bus transmission module is used to transmit the control signals based on the LIN / PWM communication bus and in a time-division multiple access manner, and to generate multi-mode coordinated control signals for the electronic water pump.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor, configured to execute the computer program to implement the steps of the multi-mode cooperative control method for an electronic water pump based on LIN / PWM communication as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, enables the implementation of the multi-mode cooperative control method for an electronic water pump based on LIN / PWM communication as described in any one of claims 1 to 7.

Images