Method, system and device for predicting sound quality of aerodynamic noise of passenger compartment and storage medium

Abstract

The invention discloses a passenger compartment aerodynamic noise sound quality prediction method, system and device and a storage medium, and the method comprises the steps: obtaining aerodynamic noise signals in different driving scenes, and generating an average sound pressure level spectrum of a vehicle window surface corresponding to the aerodynamic noise signals through numerical simulation; extracting corresponding psychological acoustic parameters according to the aerodynamic noise signals, and performing quantitative scoring on the aerodynamic noise signals to obtain corresponding sound quality evaluation values; generating a training sample according to the average sound pressure level spectrum, the psychological acoustic parameters and the sound quality evaluation value, and training a BP neural network pre-optimized based on an improved whale optimization algorithm according to the training sample to obtain an aerodynamic noise sound quality prediction model; and performing aerodynamic noise sound quality prediction on the passenger compartment of the target vehicle according to the aerodynamic noise sound quality prediction model. The passenger compartment aerodynamic noise quality prediction accuracy and stability are improved, and the method can be applied to the technical field of artificial intelligence.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method, system, device and storage medium for predicting the acoustic quality of aerodynamic noise in a passenger cabin. Background Art

[0002] In recent years, with the development of automotive vibration and noise control technology, traditional research on automotive sound pressure level can no longer fully meet consumers' demands for automotive ride comfort. More and more researchers have turned to the study of automotive acoustic quality, that is, people's subjective feelings about noise in a specific environment. Acoustic quality can be quantified through the measurement and calculation of psychoacoustic parameters. Currently, most research methods for automotive acoustic quality focus on establishing prediction models between subjective evaluation results and objective psychoacoustic parameters.

[0003] In the prior art, Zhang Jiaochao, Yoon, etc. used the multiple linear regression method to establish a fitting model between objective acoustic parameters and subjective evaluation results, and conducted acoustic quality analysis on automotive transmissions and heating, ventilation, and air conditioning systems (HAVC). Shen Xiumin, etc. used the backpropagation (BP) neural network to establish an acoustic quality prediction model. This method has a good non-linear mapping relationship, but due to the uncertainty of the initial weights and thresholds, the stability of the prediction results is poor. In addition, Sun Liwen, etc. studied the acoustic quality of in-vehicle noise under full-throttle acceleration conditions, Huang Haibo, etc. studied the acoustic quality of in-vehicle noise under constant-speed driving conditions, and Yang Yi, etc. studied the acoustic quality of in-vehicle wind vibration noise under different opening degrees of the sunroof and left and right side windows.

[0004] Although the above research has made certain progress, the existing methods still have unsatisfactory acoustic quality prediction effects under different working conditions, especially for the prediction of the acoustic quality of aerodynamic noise in complex environments. For example, the prediction of the acoustic quality of in-vehicle noise under different rain conditions still faces challenges, and the existing methods are difficult to achieve ideal prediction effects and have poor stability. In addition, the cost of traditional wind tunnel tests is high and it is difficult to carry out on a large scale. Therefore, there is an urgent need for a more effective and stable method for predicting the acoustic quality of aerodynamic noise. Summary of the Invention

[0005] The purpose of the present invention is to solve at least to a certain extent one of the technical problems existing in the prior art.

[0006] To this end, an object of an embodiment of the present invention is to provide a method for predicting the acoustic quality of aerodynamic noise in a passenger cabin, which improves the accuracy and stability of predicting the acoustic quality of aerodynamic noise in a passenger cabin.

[0007] Another object of an embodiment of the present invention is to provide a system for predicting the acoustic quality of aerodynamic noise in a passenger cabin.

[0008] To achieve the above technical objectives, the technical solutions adopted in the embodiments of the present invention include:

[0009] In a first aspect, an embodiment of the present invention provides a method for predicting the acoustic quality of aerodynamic noise in a passenger compartment, including the following steps:

[0010] Obtain aerodynamic noise signals under different driving scenarios, and generate the average sound pressure level spectrum of the corresponding window surface through numerical simulation for the aerodynamic noise signals;

[0011] Extract the corresponding psychoacoustic parameters according to the aerodynamic noise signals, and quantitatively score the aerodynamic noise signals to obtain the corresponding acoustic quality evaluation values;

[0012] Generate training samples according to the average sound pressure level spectrum, the psychoacoustic parameters, and the acoustic quality evaluation values, and train a BP neural network pre-optimized based on an improved whale optimization algorithm according to the training samples to obtain an aerodynamic noise acoustic quality prediction model;

[0013] Predict the aerodynamic noise acoustic quality of the passenger compartment of the target vehicle according to the aerodynamic noise acoustic quality prediction model.

[0014] Further, in an embodiment of the present invention, the step of obtaining aerodynamic noise signals under different driving scenarios and generating the average sound pressure level spectrum of the corresponding window surface through numerical simulation specifically includes:

[0015] Collect the aerodynamic noise signals under different rain conditions and vehicle speeds;

[0016] Simulate the unidirectional flow field without rain through the Realizable k-ε turbulence model, and add the Discrete Phase Model (DPM) to simulate the two-phase flow field with rain to establish a numerical simulation environment;

[0017] Select multiple monitoring points on the window surface, perform transient calculations on each monitoring point to obtain transient pressure pulsation signals, and calculate the time-domain pressure signals of each monitoring point through the FW-H acoustic model;

[0018] Perform windowing and fast Fourier transform on the time-domain pressure signals to obtain the pulsation pressure spectra of each monitoring point, and then determine the average sound pressure level spectrum of the window surface according to the pulsation pressure spectra.

[0019] Further, in an embodiment of the present invention, the psychoacoustic parameters include loudness, roughness, flutter, sharpness, speech intelligibility, speech interference, and sound pressure level.

[0020] Furthermore, in one embodiment of the present invention, the quantified scoring of the aerodynamic noise signal to obtain a corresponding sound quality evaluation value specifically includes:

[0021] A plurality of professionals quantitatively scores the aerodynamic noise signal using a grading method to obtain a plurality of subjective evaluation scores;

[0022] Calculating the Spearman correlation coefficient between each pair of the current subjective evaluation scores, and obtaining the average correlation coefficient of the current subjective evaluation scores based on the Spearman correlation coefficient;

[0023] When the average correlation coefficient is greater than or equal to a preset first threshold, determining the average of the current subjective evaluation scores as the sound quality evaluation value;

[0024] When the average correlation coefficient is less than the first threshold, the current subjective evaluation scores are screened out according to the Spearman correlation coefficient, and the process returns to the step of calculating the Spearman correlation coefficients between each pair of the current subjective evaluation scores.

[0025] Furthermore, in one embodiment of the present invention, the passenger compartment aerodynamic noise sound quality prediction method further includes the step of optimizing the BP neural network based on the improved whale optimization algorithm, which specifically includes:

[0026] Initialize the weight parameters and threshold parameters of the BP neural network, generate multiple corresponding individuals, and construct the initial population;

[0027] Update the position of each individual in the current population by surrounding prey, attacking with a bubble net, and searching for prey;

[0028] Calculate the fitness value of each individual after the position update, sort the individuals in descending order according to the fitness value, and select the first half of the individuals as the elite individual set;

[0029] Calculating the reverse solution of each individual in the elite individual set and determining the fitness value of the reverse solution;

[0030] Selecting the best individual and the worst individual from the elite individual set, calculating a hybrid reverse solution, and determining a fitness value of the hybrid reverse solution;

[0031] Replace the original individuals in the elite individual set with the reverse solution / mixed reverse solution whose fitness value is greater than that of the original individual to obtain the optimized current population;

[0032] When the preset convergence condition is reached, the optimal weight parameter and the optimal threshold parameter of the BP neural network are determined according to the optimal individual in the current population, and then assigned to the BP neural network.

[0033] Further, in an embodiment of the present invention, training the BP neural network pre-optimized based on the improved whale optimization algorithm according to the training samples to obtain an aerodynamic noise sound quality prediction model, which specifically includes:

[0034] Inputting the average sound pressure level spectrum and the psychoacoustic parameters into the BP neural network to obtain a sound quality prediction value;

[0035] Determining a loss value according to the sound quality prediction value and the sound quality evaluation value;

[0036] Updating the parameters of the BP neural network according to the loss value to obtain the aerodynamic noise sound quality prediction model.

[0037] Further, in an embodiment of the present invention, predicting the aerodynamic noise sound quality of the target vehicle occupant compartment according to the aerodynamic noise sound quality prediction model, which specifically includes:

[0038] Obtaining the target noise signal of the target vehicle occupant compartment and determining the target average sound pressure level spectrum of the corresponding window surface;

[0039] Extracting the corresponding target psychoacoustic parameters according to the target noise signal;

[0040] Inputting the target average sound pressure level spectrum and the target psychoacoustic parameters into the aerodynamic noise sound quality prediction model to obtain the aerodynamic noise sound quality prediction result of the target vehicle occupant compartment.

[0041] In a second aspect, an embodiment of the present invention provides an occupant compartment aerodynamic noise sound quality prediction system, including:

[0042] A data acquisition module, configured to acquire aerodynamic noise signals in different driving scenarios and generate the average sound pressure level spectrum of the corresponding window surface through numerical simulation;

[0043] A sound quality evaluation module, configured to extract the corresponding psychoacoustic parameters according to the aerodynamic noise signal and perform quantitative scoring on the aerodynamic noise signal to obtain the corresponding sound quality evaluation value;

[0044] A model training module, configured to generate training samples according to the average sound pressure level spectrum, the psychoacoustic parameters, and the sound quality evaluation value, and train the BP neural network pre-optimized based on the improved whale optimization algorithm according to the training samples to obtain an aerodynamic noise sound quality prediction model;

[0045] A sound quality prediction module, configured to predict the aerodynamic noise sound quality of the target vehicle occupant compartment according to the aerodynamic noise sound quality prediction model.

[0046] In a third aspect, an embodiment of the present invention provides an occupant compartment aerodynamic noise sound quality prediction device, including:

[0047] At least one processor;

[0048] At least one memory for storing at least one program;

[0049] When the at least one program is executed by the at least one processor, the at least one processor is caused to implement the above-mentioned occupant compartment aerodynamic noise sound quality prediction method.

[0050] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, in which a program executable by a processor is stored, and the program executable by the processor is used to execute the above-mentioned occupant compartment aerodynamic noise sound quality prediction method when executed by the processor.

[0051] The advantages and beneficial effects of the present invention will be partially given in the following description, partially will become obvious from the following description, or will be understood through the practice of the present invention:

[0052] The embodiment of the present invention acquires aerodynamic noise signals under different driving scenarios, generates the average sound pressure level spectrum of the window surface corresponding to the aerodynamic noise signals through numerical simulation, extracts the corresponding psychoacoustic parameters from the aerodynamic noise signals, quantifies and scores the aerodynamic noise signals to obtain the corresponding sound quality evaluation value, generates training samples based on the average sound pressure level spectrum, psychoacoustic parameters and sound quality evaluation value, trains a BP neural network pre-optimized based on an improved whale optimization algorithm according to the training samples to obtain an aerodynamic noise sound quality prediction model, and predicts the aerodynamic noise sound quality of the target vehicle occupant compartment according to the aerodynamic noise sound quality prediction model. The embodiment of the present invention reduces the experimental cost through real vehicle road tests and numerical simulation methods, optimizes the BP neural network using an improved whale optimization algorithm, solves the problems of poor stability and low prediction accuracy of existing methods, and improves the accuracy and stability of the prediction of the aerodynamic noise sound quality of the occupant compartment. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following introduces the drawings required to be used in the embodiments of the present invention. It should be understood that the drawings introduced below only facilitate the clear expression of some embodiments of the technical solutions in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative efforts.

[0054] Figure 1 It is a flowchart of the steps of an occupant compartment aerodynamic noise sound quality prediction method provided by an embodiment of the present invention;

[0055] Figure 2A structural block diagram of an occupant compartment aerodynamic noise sound quality prediction system provided by an embodiment of the present invention;

[0056] Figure 3 A structural block diagram of an occupant compartment aerodynamic noise sound quality prediction device provided by an embodiment of the present invention. Specific embodiments

[0057] The following describes in detail the embodiments of the present invention. The examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary and are only used to explain the present invention and should not be construed as a limitation to the present invention. For the step numbers in the following embodiments, they are only set for the convenience of elaboration and explanation, and no limitation is imposed on the order between the steps. The execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0058] In the description of the present invention, the meaning of "a plurality" is two or more. If the first and second are described, it is only for the purpose of distinguishing technical features and should not be construed as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features or implicitly indicating the sequence of the indicated technical features. In addition, unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field of the present invention.

[0059] Existing sound quality prediction methods, such as multiple linear regression and traditional BP neural networks, although they can establish the relationship between subjective evaluation results and psychoacoustic objective parameters to a certain extent, due to the uncertainty of the initial weights and thresholds, the stability of the prediction results is poor. This makes it difficult to ensure the consistency and reliability of the prediction results in practical applications. In complex environments, such as different rain conditions and vehicle speeds, existing sound quality prediction methods are difficult to achieve ideal prediction effects. These methods often ignore the dynamic characteristics of aerodynamic noise and the complex influence of environmental factors, resulting in large deviations in the prediction results. In addition, traditional wind tunnel tests are costly, difficult to carry out on a large scale, and difficult to simulate the noise characteristics under various complex environmental conditions. <00>

[0060] Refer to Figure 1 , the embodiments of the present invention provide an occupant compartment aerodynamic noise sound quality prediction method, which specifically includes the following steps:

[0061] S101. Obtain the aerodynamic noise signals under different driving scenarios, and generate the average sound pressure level spectrum of the window surface corresponding to the aerodynamic noise signals through numerical simulation;

[0062] S102. Extract the corresponding psychoacoustic parameters from the aerodynamic noise signal, quantify and score the aerodynamic noise signal to obtain the corresponding sound quality evaluation value;

[0063] S103. Generate training samples based on the average sound pressure level spectrum, psychoacoustic parameters, and sound quality evaluation value, and train the BP neural network pre-optimized by the improved whale optimization algorithm according to the training samples to obtain an aerodynamic noise sound quality prediction model;

[0064] S104. Predict the aerodynamic noise sound quality of the target vehicle occupant compartment according to the aerodynamic noise sound quality prediction model.

[0065] Specifically, in the embodiment of the present invention, through real vehicle road tests and numerical simulation methods, the experimental cost is reduced. The improved whale optimization algorithm is used to optimize the BP neural network, solving the problems of poor stability and low prediction accuracy of the existing methods, and improving the accuracy and stability of the prediction of the aerodynamic noise sound quality of the occupant compartment.

[0066] Further as an optional implementation manner, obtain the aerodynamic noise signals under different driving scenarios, and generate the average sound pressure level spectrum of the window surface corresponding to the aerodynamic noise signal through numerical simulation, which specifically includes:

[0067] S1011. Collect the aerodynamic noise signals under different rain conditions and vehicle speeds;

[0068] S1012. Simulate the unidirectional flow field without rain through the Realizable k-ε turbulence model, add the discrete phase model DPM to simulate the two-phase flow field with rain, and establish a numerical simulation environment;

[0069] S1013. Select multiple monitoring points on the window surface, perform transient calculations on each monitoring point to obtain the transient pressure pulsation signal, and calculate the time-domain pressure signal of each monitoring point through the FW-H acoustic model;

[0070] S1014. Window and perform fast Fourier transform on the time-domain pressure signal to obtain the pulsating pressure spectrum of each monitoring point, and then determine the average sound pressure level spectrum of the window surface according to the pulsating pressure spectrum.

[0071] Specifically, referring to the GB / T 18697—2002 standard, a SAVANT MI-7016 data collector and a microphone were used to collect in-vehicle noise signals under different rain conditions and vehicle speeds. The microphone was arranged at the left ear of the driver, with the vertical position being 0.70 ± 0.05 m above the intersection line of the seat surface and the headrest surface, and the horizontal position being 0.20 ± 0.02 m to the right of the longitudinal section of the seat. Noise signals were recorded at vehicle speeds of 80 km / h, 90 km / h, 100 km / h, 110 km / h, and 120 km / h under rainless, light rain (2.5 mm / h), moderate rain (8.0 mm / h), and heavy rain (16.0 mm / h) weather conditions, for a total of 20 groups. The Realizable k-ε turbulence model was used to simulate the unidirectional flow field under rainless conditions, and the discrete phase model (DPM) was added to simulate the two-phase flow field under rainy conditions. Several monitoring points were selected on the surface of the left front window, transient calculations were performed and the FW-H model was enabled, 0.5 seconds of data was taken after the calculation, the pulsating pressure spectra of each monitoring point were obtained, and the average sound pressure level spectrum of the left front window was obtained through Fourier transform.

[0072] Further as an optional implementation manner, the psychoacoustic parameters include loudness, roughness, flutter, sharpness, speech intelligibility, speech interference, and sound pressure level.

[0073] Specifically, seven psychoacoustic parameters such as loudness, roughness, flutter, sharpness, speech intelligibility, speech interference, and sound pressure level were extracted from the noise data obtained from the real vehicle road test, and these parameters will be used as the input of the subsequent BP neural network.

[0074] Further as an optional implementation manner, the aerodynamic noise signal was quantitatively scored to obtain the corresponding sound quality evaluation value, which specifically includes:

[0075] S1021. The aerodynamic noise signal was quantitatively scored by multiple professionals using the grading method to obtain multiple subjective evaluation scores;

[0076] S1022. Calculate the Spearman correlation coefficient between each pair of the current subjective evaluation scores, and obtain the average correlation coefficient of the current subjective evaluation scores according to the Spearman correlation coefficient;

[0077] S1023. When the average correlation coefficient is greater than or equal to the preset first threshold, determine the mean of the current subjective evaluation scores as the sound quality evaluation value;

[0078] S1024. When the average correlation coefficient is less than the first threshold, screen out the current subjective evaluation scores according to the Spearman correlation coefficient, and return to the step of calculating the Spearman correlation coefficient between each pair of the current subjective evaluation scores.

[0079] Specifically, the collected noise signals are quantitatively scored using a grading method, and the evaluators are 20 researchers majoring in vehicle engineering. The Spearman rank correlation coefficient is used to verify the accuracy of the subjective evaluation results to ensure the consistency and reliability of the evaluation results.

[0080] Further as an optional implementation manner, the occupant compartment aerodynamic noise sound quality prediction method further includes the step of optimizing a BP neural network based on an improved whale optimization algorithm, which specifically includes:

[0081] S201. Initialize the weight parameters and threshold parameters of the BP neural network, generate corresponding multiple individuals, and construct an initial population;

[0082] S202. Update the positions of each individual in the current population by surrounding the prey, attacking with a bubble net, and searching for prey;

[0083] S203. Calculate the fitness values of each individual after the position update, sort the individuals in descending order according to the fitness values, and select the first half of the individuals as the elite individual set;

[0084] S204. Calculate the reverse solutions of each individual in the elite individual set and determine the fitness values of the reverse solutions;

[0085] S205. Select the optimal individual and the worst individual in the elite individual set, calculate the mixed reverse solution, and determine the fitness value of the mixed reverse solution;

[0086] S206. Use the reverse solution / mixed reverse solution with a fitness value greater than the original individual to replace the original individual in the elite individual set to obtain the optimized current population;

[0087] S207. When the preset convergence condition is reached, determine the optimal weight parameters and optimal threshold parameters of the BP neural network according to the optimal individual in the current population, and then assign them to the BP neural network.

[0088] Specifically, the embodiment of the present invention improves the weight parameters and threshold parameters of the BP neural network by optimizing the Whale Optimization Algorithm (IWOA). First, the weights and thresholds of the BP neural network are initialized to obtain the initial population. Then, the parameters of the whale optimization algorithm are set, including the population size, the maximum number of iterations, etc. Next, the positions of the whale individuals are updated through three stages of surrounding prey, bubble-net attacking, and searching for prey in combination with the hybrid reverse learning strategy, gradually approaching the global optimal solution. Specifically, by combining the lens imaging reverse learning strategy and the optimal-worst reverse learning strategy, the positions of the initial population are iterated and optimized. That is, the individuals of the current population are sorted according to the fitness, and the first half of the individuals with larger fitness are taken. Their reverse populations (including the lens direction solution and the optimal-worst reverse solution) are calculated and compared with the original whale population. The individuals with better fitness are retained, and the better reverse population factors are introduced to enhance the global optimization ability of the algorithm. The optimized weights and thresholds are assigned to the BP neural network for subsequent training and testing, and compared and analyzed with the BP neural network before optimization to verify the optimization effect.

[0089] Further as an optional implementation manner, the BP neural network pre-optimized based on the improved whale optimization algorithm is trained according to the training samples to obtain an aerodynamic noise sound quality prediction model, which specifically includes:

[0090] S1031. Input the average sound pressure level spectrum and psychoacoustic parameters into the BP neural network to obtain the sound quality prediction value;

[0091] S1032. Determine the loss value according to the sound quality prediction value and the sound quality evaluation value;

[0092] S1033. Update the parameters of the BP neural network according to the loss value to obtain the aerodynamic noise sound quality prediction model.

[0093] Specifically, the optimized BP neural network is trained and tested using the training samples obtained in the foregoing steps, and the prediction results before and after optimization are compared to verify the optimization effect. Furthermore, the parameters are adjusted to optimize the model performance to ensure the accuracy and stability of the prediction results.

[0094] Further as an optional implementation manner, the aerodynamic noise sound quality of the target vehicle occupant compartment is predicted according to the aerodynamic noise sound quality prediction model, which specifically includes:

[0095] S1041. Obtain the target noise signal of the target vehicle occupant compartment and determine the target average sound pressure level spectrum of the corresponding window surface;

[0096] S1042. Extract the corresponding target psychoacoustic parameters according to the target noise signal;

[0097] S1043. Input the target average sound pressure level spectrum and the target psychoacoustic parameters into the aerodynamic noise sound quality prediction model to obtain the prediction result of the aerodynamic noise sound quality in the target vehicle occupant compartment.

[0098] Specifically, in the real vehicle scenario, the aerodynamic noise sound quality prediction model can be deployed on the vehicle end (lightweight) or roadside equipment, and the aerodynamic noise sound quality prediction is performed based on the real-time collected data, so as to obtain the prediction result of the aerodynamic noise sound quality in the target vehicle occupant compartment.

[0099] The method steps of the embodiments of the present invention are described above. It can be understood that through the real vehicle road test and numerical simulation method in the embodiments of the present invention, the experimental cost is reduced. By using the improved whale optimization algorithm to optimize the BP neural network, the problems of poor stability and low prediction accuracy of the existing methods are solved, and the accuracy and stability of the prediction of the aerodynamic noise sound quality in the occupant compartment are improved.

[0100] Compared with the prior art, the embodiments of the present invention also have the following advantages:

[0101] 1) Improve prediction stability and accuracy: The traditional BP neural network has problems such as slow learning speed, poor generalization, and instability. This solution improves the global optimization ability and stability of the algorithm by introducing the improved whale optimization algorithm (IWOA) and combining the hybrid backpropagation learning strategy. Specifically, through the lens imaging backpropagation learning strategy and the best-worst backpropagation learning strategy, the positions of the initial population are optimized, and the convergence speed and optimization ability of the algorithm are improved. This improvement makes the optimization of the weights and thresholds of the BP neural network more efficient and stable, thereby improving the overall performance of the prediction model.

[0102] 2) Efficient and economical experimental method: The traditional wind tunnel test has a high cost and is difficult to carry out on a large scale. This solution greatly reduces the experimental cost through the real vehicle road test and numerical simulation method. Specifically, the SAVANT MI-7016 data collector and microphone are used to collect the in-vehicle noise signals under different rain conditions and vehicle speeds, and the noise distribution in different environments is simulated through the numerical simulation method. This method not only improves the flexibility and efficiency of the experiment but also can more accurately capture the noise characteristics in complex environments.

[0103] 3) Multi-dimensional psychoacoustic parameter analysis: This solution uses seven psychoacoustic parameters, such as loudness, roughness, flutter, sharpness, speech intelligibility, speech interference level, and sound pressure level, as objective evaluation indicators, which can comprehensively and meticulously analyze the subjective feelings of in-vehicle noise. Through the correlation analysis of these parameters, the sound quality can be predicted more accurately. Specifically, through the extraction and analysis of psychoacoustic parameters and the combination of the grade scoring method for subjective evaluation, the consistency and reliability of the evaluation results are ensured.

[0104] 4) Wide applicability and flexibility: This solution is applicable to a variety of complex environmental conditions, such as different rain conditions and vehicle speeds, and can effectively handle various noise scenarios. In addition, by improving the whale optimization algorithm, the global optimization ability and maturity of the algorithm are improved, and the flexibility and adaptability of the algorithm are enhanced.

[0105] Refer to Figure 2 , an embodiment of the present invention provides an occupant compartment aerodynamic noise sound quality prediction system, including:

[0106] A data acquisition module, configured to acquire aerodynamic noise signals under different driving scenarios, and generate the average sound pressure level spectrum of the window surface corresponding to the aerodynamic noise signal through numerical simulation;

[0107] A sound quality evaluation module, configured to extract corresponding psychoacoustic parameters according to the aerodynamic noise signal, and quantitatively score the aerodynamic noise signal to obtain a corresponding sound quality evaluation value;

[0108] A model training module, configured to generate training samples according to the average sound pressure level spectrum, psychoacoustic parameters, and sound quality evaluation values, and train a BP neural network pre-optimized based on the improved whale optimization algorithm according to the training samples to obtain an aerodynamic noise sound quality prediction model;

[0109] A sound quality prediction module, configured to predict the aerodynamic noise sound quality of the target vehicle occupant compartment according to the aerodynamic noise sound quality prediction model.

[0110] The content in the above method embodiments is applicable to the system embodiments. The functions specifically implemented by the system embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those of the above method embodiments.

[0111] Refer to Figure 3 , an embodiment of the present invention provides an occupant compartment aerodynamic noise sound quality prediction device, including:

[0112] At least one processor;

[0113] At least one memory, configured to store at least one program;

[0114] When the above at least one program is executed by the above at least one processor, the above at least one processor implements the above method for predicting the aerodynamic noise sound quality of the occupant compartment.

[0115] The content in the above method embodiments is applicable to the device embodiments. The functions specifically implemented by the device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those of the above method embodiments.

[0116] An embodiment of the present invention further provides a computer-readable storage medium, which stores a program executable by a processor. When the program executable by the processor is executed by the processor, it is used to execute the above-mentioned method for predicting the acoustic quality of occupant compartment aerodynamic noise.

[0117] A computer-readable storage medium according to an embodiment of the present invention can execute a method for predicting the acoustic quality of occupant compartment aerodynamic noise provided by an embodiment of the method of the present invention, can execute any combination of implementation steps of the method embodiment, and has the corresponding functions and beneficial effects of the method.

[0118] An embodiment of the present invention also discloses a computer program product or a computer program. The computer program product or the computer program includes computer instructions, and the computer instructions are stored in a computer-readable storage medium. A processor of a computer device can read the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes Figure 1 the method shown.

[0119] In some alternative embodiments, the functions / operations mentioned in the block diagram may not occur in the order mentioned in the operation diagram. For example, depending on the functions / operations involved, two consecutive blocks shown may actually be executed substantially simultaneously, or the above-mentioned blocks can sometimes be executed in the reverse order. In addition, the embodiments presented and described in the flowcharts of the present invention are provided by way of example for the purpose of providing a more comprehensive understanding of the technology. The disclosed method is not limited to the operations and logical processes presented herein. Alternative embodiments are foreseeable, in which the order of various operations is changed and the sub-operations described as part of a larger operation are executed independently.

[0120] In addition, although the present invention is described in the context of functional modules, it should be understood that, unless otherwise stated to the contrary, one or more of the above functions and / or features may be integrated in a single physical device and / or software module, or one or more functions and / or features may be implemented in separate physical devices or software modules. It can also be understood that a detailed discussion of the actual implementation of each module is not necessary for understanding the present invention. More precisely, considering the attributes, functions, and internal relationships of various functional modules in the devices disclosed herein, the actual implementation of the modules will be understood within the ordinary skills of an engineer. Therefore, those skilled in the art can implement the present invention as set forth in the claims without undue experimentation. It can also be understood that the specific concepts disclosed are merely illustrative and are not intended to limit the scope of the present invention, and the scope of the present invention is determined by the full scope of the appended claims and their equivalents.

[0121] If the above functions are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods in the various embodiments of the present invention. The aforementioned storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROMs, Read-Only Memories), random access memories (RAMs, Random Access Memories), magnetic disks, or optical discs that can store program codes.

[0122] The logic and / or steps represented in the flowchart or described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch instructions from the instruction execution system, apparatus, or device and execute the instructions), or in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

[0123] More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection part with one or more wirings (electronic device), a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, a computer-readable medium can even be paper or other suitable media on which the above program can be printed, because the above program can be obtained electronically, for example, by optically scanning the paper or other media, then editing, interpreting, or processing it in other suitable ways as necessary, and then storing it in a computer memory.

[0124] It should be understood that various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), and the like.

[0125] In the above description of this specification, the description with reference to the terms "one embodiment / example", "another embodiment / example" or "certain embodiments / examples", etc. means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0126] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the claims and their equivalents.

[0127] The above has specifically described the preferred embodiments of the present invention, but the present invention is not limited to the above embodiments. Those skilled in the art can also make various equivalent deformations or substitutions without departing from the spirit of the present invention, and these equivalent deformations or substitutions are all included within the scope defined by the claims of this application.

Claims

1. A method for predicting the sound quality of aerodynamic noise in a passenger cabin, characterized in that, Including the following steps: Obtain the aerodynamic noise signals under different driving scenarios, and generate the average sound pressure level spectrum of the window surface corresponding to the aerodynamic noise signals through numerical simulation; Extract the corresponding psychoacoustic parameters according to the aerodynamic noise signals, and quantitatively score the aerodynamic noise signals to obtain the corresponding sound quality evaluation values; Generate training samples according to the average sound pressure level spectrum, the psychoacoustic parameters, and the sound quality evaluation values, and train the BP neural network pre-optimized based on the improved whale optimization algorithm according to the training samples to obtain an aerodynamic noise sound quality prediction model; Predict the aerodynamic noise sound quality of the target vehicle occupant compartment according to the aerodynamic noise sound quality prediction model.

2. The method for predicting the aeroacoustic sound quality of an occupant cabin according to claim 1, wherein The step of obtaining the aerodynamic noise signals under different driving scenarios and generating the average sound pressure level spectrum of the window surface corresponding to the aerodynamic noise signals through numerical simulation specifically includes: Collect the aerodynamic noise signals under different rain conditions and vehicle speeds; Simulate the unidirectional flow field without rain through the Realizable k-ε turbulence model, and add the Discrete Phase Model (DPM) to simulate the two-phase flow field with rain to establish a numerical simulation environment; Select multiple monitoring points on the window surface, perform transient calculations on each monitoring point to obtain transient pressure pulsation signals, and calculate the time-domain pressure signals of each monitoring point through the FW-H acoustic model; Perform windowing and fast Fourier transform on the time-domain pressure signals to obtain the pulsation pressure spectra of each monitoring point, and then determine the average sound pressure level spectrum of the window surface according to the pulsation pressure spectra.

3. A method for predicting the acoustic quality of the occupant compartment aerodynamic noise according to claim 1, characterized in that: The psychoacoustic parameters include loudness, roughness, flutter, sharpness, speech intelligibility, speech interference, and sound pressure level.

4. A method for predicting the acoustic quality of the occupant compartment aerodynamic noise according to claim 1, characterized in that The step of quantitatively scoring the aerodynamic noise signals to obtain the corresponding sound quality evaluation values specifically includes: Quantitatively score the aerodynamic noise signals by multiple professionals using the grading scoring method to obtain multiple subjective evaluation scores; Calculate the Spearman correlation coefficient between each pair of current subjective evaluation scores, and obtain the average correlation coefficient of the current subjective evaluation scores according to the Spearman correlation coefficient; When the average correlation coefficient is greater than or equal to a preset first threshold, determine the mean of the current subjective evaluation scores as the sound quality evaluation value; When the average correlation coefficient is less than the first threshold, screen out the current subjective evaluation scores according to the Spearman correlation coefficient, and return to the step of calculating the Spearman correlation coefficient between each pair of current subjective evaluation scores.

5. A method for predicting the acoustic quality of the occupant compartment aerodynamic noise according to claim 1, characterized in that, The method for predicting the aerodynamic noise sound quality of the occupant compartment further includes the step of optimizing the BP neural network based on the improved whale optimization algorithm, which specifically includes: Initialize the weight parameters and threshold parameters of the BP neural network, generate corresponding multiple individuals, and construct an initial population; Update the positions of each individual in the current population by surrounding the prey, attacking with a bubble net, and searching for prey; Calculate the fitness values of each individual after the position update, sort the individuals in descending order according to the fitness values, and select the first half of the individuals as the elite individual set; Calculate the reverse solutions of each individual in the elite individual set, and determine the fitness values of the reverse solutions; Select the optimal individual and the worst individual in the elite individual set, calculate the hybrid reverse solution, and determine the fitness value of the hybrid reverse solution; Use the reverse solution / hybrid reverse solution with a fitness value greater than that of the original individual to replace the original individual in the elite individual set to obtain the optimized current population; When the preset convergence condition is reached, determine the optimal weight parameters and optimal threshold parameters of the BP neural network according to the optimal individual in the current population, and then assign them to the BP neural network.

6. A method for predicting the acoustic quality of the occupant compartment aerodynamic noise according to claim 1, characterized in that, Training the BP neural network pre-optimized by the improved whale optimization algorithm according to the training samples to obtain an aerodynamic noise sound quality prediction model, which specifically includes: Input the average sound pressure level spectrum and the psychoacoustic parameters into the BP neural network to obtain a sound quality prediction value; Determine the loss value according to the sound quality prediction value and the sound quality evaluation value; Update the parameters of the BP neural network according to the loss value to obtain the aerodynamic noise sound quality prediction model.

7. A method for predicting the acoustic quality of the occupant compartment aerodynamic noise according to any one of claims 1 to 6, characterized in that, Predicting the aerodynamic noise sound quality of the target vehicle occupant compartment according to the aerodynamic noise sound quality prediction model, which specifically includes: Obtain the target noise signal of the target vehicle occupant compartment, and determine the target average sound pressure level spectrum of the corresponding window surface; Extract the corresponding target psychoacoustic parameters according to the target noise signal; Input the target average sound pressure level spectrum and the target psychoacoustic parameters into the aerodynamic noise sound quality prediction model to obtain the aerodynamic noise sound quality prediction result of the target vehicle occupant compartment.

8. An occupant compartment aerodynamic noise sound quality prediction system, characterized in that, Including: A data acquisition module, configured to acquire aerodynamic noise signals under different driving scenarios, and generate the average sound pressure level spectrum of the corresponding window surface of the aerodynamic noise signal through numerical simulation; A sound quality evaluation module, configured to extract the corresponding psychoacoustic parameters according to the aerodynamic noise signal, and quantitatively score the aerodynamic noise signal to obtain the corresponding sound quality evaluation value; A model training module, configured to generate training samples according to the average sound pressure level spectrum, the psychoacoustic parameters, and the sound quality evaluation value, and train a BP neural network pre-optimized by the improved whale optimization algorithm according to the training samples to obtain an aerodynamic noise sound quality prediction model; A sound quality prediction module, configured to predict the aerodynamic noise sound quality of the target vehicle occupant compartment according to the aerodynamic noise sound quality prediction model.

9. An occupant cabin aerodynamic noise sound quality prediction device, characterized in that, Including: At least one processor; At least one memory, configured to store at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements a method for predicting the aerodynamic noise sound quality of a vehicle occupant compartment according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a program executable by a processor, characterized in that, The program executable by the processor, when executed by the processor, is used to execute a method for predicting the aerodynamic noise sound quality of a vehicle occupant compartment according to any one of claims 1 to 7.

Images