Multi-dimensional automobile sound quality evaluation method, device, equipment and storage medium

Abstract

The invention discloses a multi-dimensional automobile sound quality evaluation method, device and equipment and a storage medium, and relates to the technical field of automobile sound quality evaluation, and the method comprises the steps: obtaining a driving condition ratio of a target automobile type, in-automobile noise data under a driving condition, and a noise level psychological expectation score; based on the in-vehicle noise data, extracting acoustic features of multiple dimensions corresponding to road noise, wind noise and power assembly noise; and inputting the driving condition ratio, the acoustic features of multiple dimensions and the noise level psychological expectation score into a sound quality evaluation model to obtain a sound quality subjective evaluation score of the target vehicle model, the technical problem of large deviation between the evaluation result and the subjective feeling of the user due to single dimension and separation from the specific driving scene in the existing sound quality evaluation method is solved. The technical effect of carrying out automatic and multi-dimensional comprehensive evaluation on the vehicle sound quality in a limited scene is achieved, and the evaluation result better fits the real auditory experience of a user under different driving conditions.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of automobile sound quality evaluation, and in particular to a multi-dimensional automobile sound quality evaluation method, device, equipment and storage medium. BACKGROUND

[0002] In the objective evaluation of automobile sound quality, a method is needed that can accurately and automatically evaluate the sound quality of in-vehicle noise in a specific driving scenario. This method should effectively integrate the characteristics of different noise types and output evaluation results consistent with human subjective perception.

[0003] Existing technologies usually use a single psychoacoustic index (such as loudness, sharpness) or a simple weighted sound pressure level to evaluate in-vehicle noise. However, these indicators each only reflect a certain dimension of noise characteristics, and do not differentiate between different driving conditions and weight different noise types, resulting in evaluation results that deviate from actual subjective experiences in multiple scenarios and cannot fully and accurately represent the overall sound quality of a vehicle.

[0004] Therefore, the problem to be solved at present is how to automatically evaluate multiple types of in-vehicle noise in a specific driving scenario.

[0005] The above content is only used to assist in understanding the technical solutions of the present application and does not represent an admission that the above content is prior art. SUMMARY

[0006] The main purpose of the present application is to provide a multi-dimensional automobile sound quality evaluation method, device, equipment and storage medium, aiming to solve the technical problem of how to automatically evaluate multiple types of in-vehicle noise in a specific driving scenario.

[0007] To achieve the above purpose, the present application provides a multi-dimensional automobile sound quality evaluation method, which comprises: obtaining the driving condition proportion of a target vehicle model, in-vehicle noise data under the driving condition, and noise level psychological expectation deduction points; extracting multiple dimensions of acoustic features corresponding to road noise, wind noise and powertrain noise based on the in-vehicle noise data; inputting the driving condition proportion, the multiple dimensions of acoustic features and the noise level psychological expectation deduction points into a sound quality evaluation model to obtain a sound quality subjective evaluation score of the target vehicle model.

[0008] In an embodiment, obtaining the driving condition proportion of a target vehicle model comprises: obtaining vehicle driving data of a target vehicle model; determine, based on the vehicle driving data, an urban working condition proportion, a suburban working condition proportion, a high-speed working condition proportion, and a low-speed accelerating working condition proportion of the target vehicle model; use the urban working condition proportion, the suburban working condition proportion, the high-speed working condition proportion, and the low-speed accelerating working condition proportion as the driving working condition proportions.

[0009] In an embodiment, the driving working conditions include a road noise working condition, an air noise working condition, and a powertrain noise working condition. obtain in-vehicle noise data of the target vehicle model under preset driving working conditions, including: obtain first noise data of the target vehicle model when driving at a first preset vehicle speed on a smooth asphalt road and a rough road, and use the first noise data as in-vehicle noise data of the road noise working condition; obtain second noise data of the target vehicle model when driving at a second preset vehicle speed on a smooth asphalt road, and use the second noise data as in-vehicle noise data of the air noise working condition and the powertrain noise working condition, wherein the second preset vehicle speed is higher than the first preset vehicle speed.

[0010] In an embodiment, obtaining a noise level psychological expectation deduction of the target vehicle model includes: obtain a price range of the target vehicle model; obtain a first actual sound pressure level and a second actual sound pressure level based on the in-vehicle noise data; obtain a first preset standard sound pressure level and a second preset standard sound pressure level based on the driving working conditions corresponding to the in-vehicle noise data and the price range; obtain a noise level psychological expectation deduction based on the first actual sound pressure level, the second actual sound pressure level, the first preset standard sound pressure level, and the second preset standard sound pressure level.

[0011] In an embodiment, based on the in-vehicle noise data, acoustic characteristics of multiple dimensions corresponding to road noise, air noise, and powertrain noise are extracted, including: obtain a powertrain noise extraction vehicle speed interval; extract a total level sound pressure level, a total level loudness, a total level speech intelligibility, a total level sharpness, sound pressure level RMS values of different frequency bands, and spectral values as acoustic characteristics corresponding to road noise based on the in-vehicle noise data; extract a total level sound pressure level, a total level loudness, a total level speech intelligibility, a total level sharpness, and spectral values as acoustic characteristics corresponding to air noise based on the in-vehicle noise data; divide the in-vehicle noise data based on the powertrain noise extraction vehicle speed interval to obtain in-vehicle noise slice data at different vehicle speeds; Based on the in-vehicle noise slice data, sound pressure levels corresponding to different orders of motion, sound pressure level change rates, reducer order noise prominence ratios, motor order noise prominence ratios, sound pressure level RMS values of preset frequency bands, and frequency spectrum values are extracted as acoustic characteristics corresponding to powertrain noise.

[0012] In an embodiment, before the step of inputting the driving condition proportion, the multi-dimensional acoustic characteristics, and the noise level psychological expectation deduction score into the sound quality evaluation model to obtain the sound quality subjective evaluation score of the target vehicle, the method further comprises: obtaining driving condition historical proportions of different vehicles, historical in-vehicle noise data under driving conditions, and historical noise level psychological expectation deduction scores; Based on the historical in-vehicle noise data, extracting historical acoustic characteristics corresponding to the driving conditions; obtaining subjective evaluation scores of the target user group on the historical noise data, and taking the subjective evaluation scores as training labels; training an initial model based on the driving condition historical proportions, the historical acoustic characteristics, the historical noise level psychological expectation deduction scores, and the training labels to obtain a sound quality evaluation model.

[0013] In an embodiment, the training of the initial model based on the driving condition historical proportions, the historical acoustic characteristics, the historical noise level psychological expectation deduction scores, and the training labels to obtain a sound quality evaluation model comprises: inputting the historical acoustic characteristics into the initial model to obtain road noise scores, wind noise scores, and total motion noise scores; obtaining initial sound quality subjective evaluation scores based on the road noise scores, the wind noise scores, the total motion noise scores, the driving condition historical proportions, and the historical noise level psychological expectations; comparing the initial sound quality subjective evaluation scores with the training labels to obtain a model fitting rate; when the model fitting rate reaches a preset threshold, determining a sound quality evaluation model.

[0014] In addition, to achieve the above-mentioned purpose, the application further provides a multi-dimensional automobile sound quality evaluation device, which comprises: a data input module for obtaining driving condition proportions of a target vehicle, in-vehicle noise data under driving conditions, and noise level psychological expectation deduction scores; a feature extraction module for extracting multi-dimensional acoustic characteristics corresponding to road noise, wind noise, and powertrain noise, respectively, based on the in-vehicle noise data; The model calculation module is used to input the driving condition ratio, the acoustic features of the multiple dimensions, and the psychological expectation deduction of the noise level into the sound quality evaluation model to obtain the subjective sound quality evaluation score of the target vehicle model.

[0015] In addition, to achieve the above objectives, this application also proposes a multi-dimensional vehicle sound quality evaluation device, the device comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the multi-dimensional vehicle sound quality evaluation method as described above.

[0016] In addition, to achieve the above objectives, this application also proposes a storage medium, which is a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it implements the steps of the multi-dimensional automotive sound quality evaluation method described above.

[0017] In addition, to achieve the above objectives, this application also provides a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the multidimensional automotive sound quality evaluation method described above.

[0018] This application obtains the driving condition ratio of the target vehicle model, in-vehicle noise data under driving conditions, and psychological expectation deductions for noise levels. Based on the in-vehicle noise data, it extracts acoustic features of multiple dimensions corresponding to road noise, wind noise, and powertrain noise. The driving condition ratio, the multiple acoustic features, and the psychological expectation deductions for noise levels are input into a sound quality evaluation model to obtain the subjective sound quality evaluation score of the target vehicle model. By employing a technical approach that integrates big data on vehicle driving conditions, multi-dimensional acoustic features, and user psychological expectations, this application solves the technical problem of existing sound quality evaluation methods, which suffer from large discrepancies between evaluation results and user subjective feelings due to their single-dimensionality and detachment from specific driving scenarios. Compared with existing technologies, this application achieves automated, multi-dimensional comprehensive evaluation of vehicle sound quality under limited scenarios, making the evaluation results more closely reflect the user's real auditory experience under different driving conditions. Attached Figure Description

[0019] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0020] To more clearly illustrate the technical solutions in 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, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart illustrating an embodiment of the multidimensional automotive sound quality evaluation method of this application. Figure 2 This is a flowchart illustrating Embodiment 2 of the multidimensional automotive sound quality evaluation method of this application. Figure 3 This is a schematic diagram of the module structure of the multi-dimensional automotive sound quality evaluation device according to an embodiment of this application; Figure 4 This is a schematic diagram of the equipment structure of the hardware operating environment involved in the multi-dimensional vehicle sound quality evaluation method in this application embodiment.

[0022] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0023] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0024] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0025] The main solution of this application embodiment is: to obtain the driving condition ratio of the target vehicle, the in-vehicle noise data under the driving condition, and the psychological expectation deduction of the noise level; based on the in-vehicle noise data, to extract acoustic features of multiple dimensions corresponding to road noise, wind noise, and powertrain noise respectively; to input the driving condition ratio, the acoustic features of multiple dimensions, and the psychological expectation deduction of the noise level into the sound quality evaluation model to obtain the subjective sound quality evaluation score of the target vehicle.

[0026] In this embodiment, for ease of description, the following description uses a computer as the execution subject.

[0027] In-vehicle noise is typically evaluated using a single psychoacoustic metric (such as loudness or sharpness) or a simple weighted sound pressure level. However, these metrics only reflect one dimension of noise characteristics and do not differentiate or assign weights to the dominant noise types under different driving conditions. This leads to discrepancies between the evaluation results and actual subjective experiences in various scenarios, making it impossible to comprehensively and accurately characterize the overall sound quality of the vehicle.

[0028] This application provides a solution that integrates big data on vehicle driving conditions, multi-dimensional acoustic characteristics, and user psychological expectations. This addresses the technical problem of existing sound quality evaluation methods, which suffer from a large discrepancy between evaluation results and user subjective experiences due to their single-dimensional approach and detachment from specific driving scenarios. Compared to existing technologies, this solution achieves automated, multi-dimensional comprehensive evaluation of vehicle sound quality within defined scenarios, making the evaluation results more closely aligned with users' actual auditory experiences under different driving conditions.

[0029] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device or computer capable of performing the above functions. The following description uses a computer as an example to illustrate this embodiment and the subsequent embodiments.

[0030] Based on this, the embodiments of this application provide a multi-dimensional vehicle sound quality evaluation method, referring to... Figure 1 , Figure 1 This is a flowchart illustrating the first embodiment of the multidimensional automotive sound quality evaluation method of this application.

[0031] In this embodiment, the multi-dimensional vehicle sound quality evaluation method includes steps S10 to S30: Step S10: Obtain the driving condition percentage of the target vehicle model, the in-vehicle noise data under the driving conditions, and the psychological expected deduction for noise level. It should be noted that the target vehicle model refers to the specific vehicle model to be evaluated for sound quality; the driving condition percentage refers to the distribution ratio of different driving scenarios in the total vehicle usage time; driving condition refers to the typical operating state divided by vehicle speed and road surface type (smooth asphalt road and rough road); in-vehicle noise data refers to the sound pressure signal recorded by the acoustic acquisition equipment installed in the vehicle; the noise level psychological expectation deduction is a correction value calculated based on the difference between the user's expectation of quietness corresponding to the vehicle price and the measured noise level.

[0032] It is understandable that, due to the systematic differences in driving habits and vehicle usage scenarios among different user groups, directly using noise data under fixed operating conditions cannot accurately reflect the acoustic performance in real-world usage environments. Therefore, step S10 is performed to avoid the evaluation conditions from being out of touch with the actual user scenarios, thereby improving the practicality and accuracy of the sound quality evaluation results.

[0033] In one feasible implementation, step S10, obtaining the driving condition percentage of the target vehicle model, includes: obtaining vehicle driving data of the target vehicle model; determining the urban driving condition percentage, suburban driving condition percentage, highway driving condition percentage, and low-speed acceleration driving condition percentage of the target vehicle model based on the vehicle driving data; and using the urban driving condition percentage, the suburban driving condition percentage, the highway driving condition percentage, and the low-speed acceleration driving condition percentage as the driving condition percentage.

[0034] It should be noted that vehicle driving data refers to vehicle speed, acceleration, and geographical location collected through in-vehicle terminals or big data platforms; urban driving condition ratio refers to the proportion of time a vehicle spends driving at low speeds on urban roads; suburban driving condition ratio refers to the proportion of time a vehicle spends driving at medium speeds on suburban roads; highway driving condition ratio refers to the proportion of time a vehicle spends driving at high speeds on highways; and low-speed acceleration driving condition ratio refers to the proportion of time a vehicle spends accelerating at high throttle at low speeds.

[0035] For example, by analyzing vehicle driving data over 30 consecutive days, when vehicles are driving on urban roads at speeds ≤60 km / h... And the vehicle acceleration is <1.5 When the vehicle is traveling on suburban roads at a speed of ≤60 km / h, the driving conditions are determined to be urban conditions. And the vehicle acceleration is <1.5 When the vehicle speed is greater than 60 km / h, it is determined to be a suburban driving condition; And the vehicle acceleration is <1.5 This is determined to be a high-speed operating condition; when the vehicle acceleration > 1.5... When the condition is determined to be low-speed acceleration, according to the above standards, the urban driving condition accounts for 42%, the suburban driving condition accounts for 28%, the highway driving condition accounts for 25%, and the low-speed acceleration condition accounts for 5%.

[0036] In this embodiment, by quantifying the distribution characteristics of users' real driving scenarios, the problem of strong subjectivity in setting the weight of driving conditions in traditional evaluation methods is solved, and data support is provided for the weight allocation of different noise types.

[0037] The above are merely feasible implementations of step S10 provided in this embodiment. This embodiment does not specifically limit the specific implementation of step S10.

[0038] In one feasible implementation, the driving conditions in step S10 include road noise conditions, wind noise conditions, and powertrain noise conditions; obtaining in-vehicle noise data of the target vehicle under preset driving conditions includes: obtaining first noise data of the target vehicle when it is traveling at a first preset speed on a smooth asphalt road and a rough road, and using the first noise data as the in-vehicle noise data of the road noise condition; obtaining second noise data of the target vehicle when it is traveling at a second preset speed on a smooth asphalt road, and using the second noise data as the in-vehicle noise data of the wind noise condition and powertrain noise condition, wherein the second preset speed is higher than the first preset speed.

[0039] It should be noted that road noise conditions refer to driving conditions where noise is generated by the interaction between the tires and the road surface; wind noise conditions refer to driving conditions where noise is generated by the interaction between airflow and the vehicle body surface; powertrain noise conditions refer to driving conditions where noise is generated by the operation of the powertrain system; the first preset speed refers to a typical urban cruising speed of 60 km / h; the first noise data refers to acoustic signals that include road excitation characteristics; the second preset speed refers to a typical high-speed cruising speed of 120 km / h; the second noise data includes acoustic signals that include aerodynamic and powertrain characteristics.

[0040] For example, the in-vehicle noise data of a vehicle traveling at a constant speed of 60 km / h on a smooth asphalt road and a rough road are selected as the in-vehicle noise data for the road noise condition, and the in-vehicle noise data of a vehicle traveling at a constant speed of 120 km / h on a smooth asphalt road are selected as the in-vehicle noise data for the wind noise condition and the total power noise condition.

[0041] In this embodiment, by establishing the correspondence between typical driving scenarios and major noise sources, the problem of various sound sources being mixed together during noise feature extraction is solved, providing a clean data foundation for subsequent feature analysis.

[0042] In one feasible implementation, step S10, obtaining the psychological expectation deduction for the noise level of the target vehicle, includes: obtaining the price range of the target vehicle; obtaining a first actual sound pressure level and a second actual sound pressure level based on the in-vehicle noise data; obtaining a first preset standard sound pressure level and a second preset standard sound pressure level based on the driving conditions corresponding to the in-vehicle noise data and the price range; and obtaining the psychological expectation deduction for the noise level based on the first actual sound pressure level, the second actual sound pressure level, the first preset standard sound pressure level, and the second preset standard sound pressure level.

[0043] It should be noted that the price range refers to the market price segment of the vehicle; the first actual sound pressure level refers to the A-weighted sound pressure level when driving at a constant speed of 60 km / h on a smooth asphalt road; the second actual sound pressure level refers to the A-weighted sound pressure level when driving at a constant speed of 120 km / h on a smooth asphalt road; the first preset standard sound pressure level refers to the expected sound pressure level under the current price range at 60 km / h; the second preset standard sound pressure level refers to the expected sound pressure level under the current price range at 120 km / h.

[0044] For example, for the price range of 50,000 to 100,000 yuan, the first preset standard sound pressure level is 60dB and the second preset standard sound pressure level is 70dB; for the price range of 100,000 to 150,000 yuan, the first preset standard sound pressure level is 59dB and the second preset standard sound pressure level is 69dB; for the price range of 150,000 to 200,000 yuan, the first preset standard sound pressure level is 58dB and the second preset standard sound pressure level is 67dB; and for the price range of 200,000 to 400,000 yuan, the first preset standard sound pressure level is 56dB and the second preset standard sound pressure level is 66dB.

[0045] Users of higher-priced vehicles have higher expectations for in-vehicle noise levels. If the in-vehicle noise level is lower than the average level for vehicles in that price range, points will be deducted. ① No points will be deducted for vehicles priced below 60,000 yuan; ② Deduction points for vehicles priced between 60,000 and 100,000 yuan = (60 - A-weighted sound pressure level in the driver's inner ear on a smooth asphalt road at a constant speed of 60 km / h) * 0.5 + (70 - A-weighted sound pressure level in the driver's inner ear on a smooth asphalt road at a constant speed of 120 km / h). ③ Deduction points for vehicles priced between 100,000 and 150,000 yuan = (59 - A-weighted sound pressure level in the driver's inner ear on a smooth asphalt road at a constant speed of 60 km / h) * 0.5 + (69 - A-weighted sound pressure level in the driver's inner ear on a smooth asphalt road at a constant speed of 120 km / h). ④ Deduction points for vehicles priced between 150,000 and 200,000 yuan = (58 - A-weighted sound pressure level in the driver's inner ear on a smooth asphalt road at a constant speed of 60 km / h) * 0.5 + (67 - A-weighted sound pressure level in the driver's inner ear on a smooth asphalt road at a constant speed of 120 km / h). ⑤ Deduction points for vehicles priced between 200,000 and 400,000 yuan = (56 - A-weighted sound pressure level in the driver's inner ear on a smooth asphalt road at a constant speed of 60 km / h) * 0.5 + (66 - A-weighted sound pressure level in the driver's inner ear on a smooth asphalt road at a constant speed of 120 km / h).

[0046] In this embodiment, by introducing noise expectation standards corresponding to market positioning, the technical problem of neglecting users' psychological expectations in traditional sound quality evaluation is solved, making the evaluation results more in line with market perception.

[0047] The above are merely feasible implementations of step S10 provided in this embodiment. This embodiment does not specifically limit the specific implementation of step S10.

[0048] Step S20: Based on the in-vehicle noise data, extract acoustic features of multiple dimensions corresponding to road noise, wind noise and powertrain noise respectively. It should be noted that the acoustic characteristics corresponding to road noise are quantitative parameters reflecting the low-to-mid frequency vibration noise generated by the interaction between the tire and the road surface; the acoustic characteristics corresponding to wind noise are quantitative parameters characterizing the mid-to-high frequency turbulent noise generated by the interaction between airflow and the vehicle body structure; and the acoustic characteristics corresponding to powertrain noise are quantitative parameters describing the characteristic order noise and broadband noise generated by the operation of power systems such as the engine, transmission, and motor.

[0049] It is understandable that, since there are fundamental differences in the acoustic mechanisms and perception characteristics of different types of noise, using a uniform feature extraction method would result in the loss of key acoustic information. Therefore, performing step S20 can avoid the distortion of the evaluation model due to insufficient feature representation capabilities, thereby improving the accuracy and discriminative power of sound quality analysis.

[0050] In one feasible implementation, step S20 may include: obtaining the powertrain noise extraction speed interval; based on the in-vehicle noise data, extracting the overall sound pressure level, overall loudness, overall speech intelligibility, overall sharpness, sound pressure level RMS value and spectral value of different frequency bands as acoustic features corresponding to road noise; based on the in-vehicle noise data, extracting the overall sound pressure level, overall loudness, overall speech intelligibility, overall sharpness and spectral value as acoustic features corresponding to wind noise; dividing the in-vehicle noise data based on the powertrain noise extraction speed interval to obtain in-vehicle noise slice data at different vehicle speeds; based on the in-vehicle noise slice data, extracting the sound pressure level, sound pressure level change rate, reducer order noise prominence ratio, motor order noise prominence ratio, sound pressure level RMS value and spectral value of preset frequency bands as acoustic features corresponding to powertrain noise for different powertrain orders.

[0051] It should be noted that total sound pressure level (SPL) can be the sum of the sound pressure energy of all frequency components across the entire audible frequency range, expressed in decibels (dB); total loudness can be calculated by performing A-weighted filtering on the original, flat noise signal to attenuate low-frequency energy that, although physically present, is insensitive to the human ear, and then calculating the total loudness value of the A-weighted signal; total speech intelligibility can be measured using the AI ​​index; total sharpness is a psychoacoustic parameter characterizing the proportion of high-frequency components; and so on. The sound pressure level RMS value in the same frequency band can include the infrasound sound pressure level RMS value of 0-20Hz, the drum sound sound pressure level RMS value of 20-60Hz, the boom sound sound pressure level RMS value of 50-120Hz, the rumble sound sound pressure level RMS value of 120-400Hz, and the tire noise sound pressure level RMS value of greater than 400Hz. The spectrum value can be a set of values ​​representing the sound energy in each narrow frequency band, obtained by using a 1 / 3 octave band analysis method within the human audible frequency range of 0 to 20000 Hz. Additionally, it should be noted that the powertrain noise extraction speed interval is an indicator used to segment continuous in-vehicle noise data according to vehicle speed. For example, when the vehicle speed increases by 5 km / h, acoustic features are extracted based on the in-vehicle noise slice data at that time. In-vehicle noise slice data refers to a series of stable slice data at different vehicle speeds obtained by dividing the original in-vehicle noise data based on the above interval. The sound pressure level corresponding to different powertrain orders refers to the sound pressure level of characteristic frequency components related to the harmonics of engine speed or motor speed, which can include the second-order, fourth-order, sixth-order, and eighth-order noise sound pressure levels of the powertrain. The sound pressure level change rate refers to the acceleration process. The slope of the sound pressure level of each order as a function of vehicle speed; the noise prominence ratio of the reducer order refers to the difference between the characteristic order sound pressure level of the reducer and the background noise sound pressure level; the noise prominence ratio of the motor order refers to the difference between the electromagnetic order sound pressure level of the motor and the background noise sound pressure level; the RMS value of the sound pressure level in the preset frequency band can be the RMS value of the sound pressure level in the range of 200-800Hz. Specifically, when extracting the acoustic features of powertrain noise in multiple dimensions, it is necessary to extract the sound pressure level, sound pressure level change rate, reducer noise prominence ratio, motor noise prominence ratio, RMS value of the sound pressure level in the preset frequency band, and spectrum value corresponding to different powertrain orders from the in-vehicle noise slice data for every 5km / h increase in vehicle speed during vehicle acceleration.

[0052] In this embodiment, by designing differentiated feature extraction schemes for the acoustic characteristics of different noise sources, the technical problems of insufficient feature set specificity and inadequate representation in traditional methods are solved, providing complete feature input for subsequent comprehensive evaluation.

[0053] The above are merely feasible implementations of step S20 provided in this embodiment. This embodiment does not specifically limit the specific implementation of step S20.

[0054] Step S30: Input the driving condition ratio, the acoustic features of the multiple dimensions, and the psychological expectation deduction of the noise level into the sound quality evaluation model to obtain the subjective sound quality evaluation score of the target vehicle model. It should be noted that the sound quality evaluation model refers to a mathematical model trained using machine learning algorithms that can simulate human subjective auditory perception; the subjective sound quality evaluation score refers to the standardized evaluation value output by the model, used to quantify the sound quality level of the target vehicle model, usually expressed on a ten-point scale. The user experience corresponding to 1-10 is shown in the table below:

[0055] The formula for calculating the subjective evaluation score of sound quality is as follows: Vehicle Noise Quality Score = Driver's Inner Ear Road Noise Score on Smooth Asphalt Road at Constant Speed ​​of 60km / h × Urban Driving Condition Percentage × Smooth Asphalt Road Coefficient + Driver's Inner Ear Road Noise Score on Rough Road at Constant Speed ​​of 60km / h × Suburban Driving Condition Percentage × Rough Road Coefficient + Driver's Inner and Outer Ear Wind Noise Score × High-Speed ​​Driving Condition Percentage + Driver's Inner Ear Noise Score for 0-100kph Acceleration × Low-Speed ​​Acceleration Driving Condition Percentage - Psychological Expectation Deduction for Vehicle Noise Level

[0056] Among them, the inner ear road noise score of the driver on a smooth asphalt road at a constant speed of 60km / h and the inner ear road noise score of the driver on a rough road at a constant speed of 60km / h are calculated based on the acoustic characteristics of multiple dimensions corresponding to road noise; the inner and outer ear wind noise score of the driver is calculated based on the acoustic characteristics of multiple dimensions corresponding to wind noise; and the inner ear noise score of the driver during acceleration from 0-100kph is calculated based on the acoustic characteristics of multiple dimensions corresponding to powertrain noise.

[0057] Understandably, since sound quality is a complex perceptual phenomenon involving multi-dimensional acoustic characteristics, usage scenarios, and psychological expectations, traditional single-parameter or linear weighting methods are difficult to accurately fit its nonlinear relationship. Therefore, step S30 can avoid the subjectivity and experience limitations of manually setting weights, thereby improving the objectivity, repeatability, and consistency with real user perception of the evaluation results.

[0058] This embodiment provides a multi-dimensional vehicle sound quality evaluation method. It obtains the driving condition percentage of a target vehicle model, in-vehicle noise data under driving conditions, and psychological expectation deductions for noise levels. Based on the in-vehicle noise data, it extracts acoustic features corresponding to road noise, wind noise, and powertrain noise in multiple dimensions. The driving condition percentage, the multiple acoustic features, and the psychological expectation deductions for noise levels are input into a sound quality evaluation model to obtain the subjective sound quality evaluation score of the target vehicle model. By employing a technique that integrates big data on vehicle driving conditions, multi-dimensional acoustic features, and user psychological expectations, it solves the technical problem of existing sound quality evaluation methods, which suffer from large discrepancies between evaluation results and user subjective feelings due to their single-dimensionality and detachment from specific driving scenarios. Compared with existing technologies, this method achieves automated, multi-dimensional comprehensive evaluation of vehicle sound quality under limited scenarios, making the evaluation results more closely reflect the user's real auditory experience under different driving conditions.

[0059] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in the first embodiment described above can be referred to the above description, and will not be repeated hereafter. Based on this, please refer to... Figure 2 Before step S30, the multi-dimensional vehicle sound quality evaluation method further includes steps S301 to S304: Step S301: Obtain the historical percentage of driving conditions for different vehicles, historical in-vehicle noise data under driving conditions, and the expected deduction for historical noise levels. It should be noted that the historical driving condition percentage of different vehicles refers to the time distribution ratio of various driving scenarios obtained from the actual use of multiple marketed models; historical in-vehicle noise data refers to the acoustic sample set of existing models collected under standard test conditions; and the psychological expectation deduction for historical noise level refers to the expected deviation value calculated based on the correspondence between the historical selling price and noise level of existing models.

[0060] It is understandable that since data from a single vehicle model is insufficient to cover the diversity of scenarios and the completeness of features required for sound quality evaluation, step S301 is performed to avoid the model overfitting problem caused by a single training data set, thereby improving the model's generalization ability across different vehicle models and scenarios.

[0061] Step S302: Based on the historical in-vehicle noise data, extract the historical acoustic features corresponding to the driving conditions; It should be noted that historical acoustic characteristics refer to a set of multi-dimensional acoustic parameters related to road noise, wind noise, and powertrain noise, calculated from historical noise data.

[0062] It is understandable that since the original noise data contains a lot of redundant information and cannot directly represent the characteristics of human auditory perception, performing step S302 can avoid the problems of high computational complexity and insignificant features when the model directly processes the original sound signal, thereby improving the model training efficiency and feature representation ability.

[0063] Step S303: Obtain the subjective evaluation scores of the target user group on the historical noise data, and use the subjective evaluation scores as training labels; It should be noted that the subjective evaluation score refers to the standardized rating value given by the target user group to historical noise samples through auditory experiments; the training label refers to the target output value used for supervised learning in the machine learning model.

[0064] It is understandable that, since sound quality evaluation is essentially based on human subjective perception as the ultimate standard, the lack of real user evaluation data will lead to the model optimization target being inaccurate. Therefore, performing step S303 can prevent the model learning target from deviating from the real perception requirements, thereby improving the consistency between the model output results and human subjective feelings.

[0065] Step S304: Train the initial model based on the historical proportion of driving conditions, the historical acoustic features, the psychological expectation deduction of historical noise levels, and the training labels to obtain a sound quality evaluation model.

[0066] It should be noted that the initial model refers to the untrained machine learning model architecture, whose network parameters are in a state of random initialization.

[0067] Understandably, since untrained models lack the ability to map multi-dimensional features to subjective scores, step S304 avoids directly using unverified mathematical models for evaluation, thereby improving the reliability and accuracy of the final sound quality evaluation results.

[0068] In one feasible implementation, step S304 may include: inputting the historical acoustic features into an initial model to obtain a road noise score, a wind noise score, and a total dynamic noise score; obtaining an initial subjective sound quality evaluation score based on the road noise score, the wind noise score, the total dynamic noise score, the historical proportion of the driving condition, and the psychological expectation of the historical noise level; comparing the initial subjective sound quality evaluation score with the training labels to obtain a model fitting rate; and determining the sound quality evaluation model when the model fitting rate reaches a preset threshold.

[0069] It should be noted that the road noise score is a sub-score calculated by the model based on the acoustic features related to road noise; the wind noise score is a sub-score calculated by the model based on the acoustic features related to wind noise; the powertrain noise score is a sub-score calculated by the model based on the acoustic features related to powertrain noise; the initial subjective sound quality evaluation score is the unoptimized total score obtained by combining the sub-scores and correction terms; the model fit rate refers to the statistical consistency index between the model's predicted score and the training labels; and the preset threshold refers to the pre-set standard for model training qualification.

[0070] For example, a linear regression model is used as the initial model. The acoustic characteristics of road noise, wind noise, and total dynamic noise are input into three sub-models respectively. After obtaining the scores of each module, they are weighted according to the proportion of urban conditions (30%) and highway conditions (20%), and then the psychological expectation deduction for historical noise levels is subtracted to obtain the initial subjective evaluation score of sound quality. When the correlation coefficient between the model's predicted score and the average score of the human scoring group reaches 0.95, it is determined that the preset threshold has been reached.

[0071] In this embodiment, a multi-stage optimization strategy combining modular training and overall fitting is adopted to solve the technical problem that a single model cannot accurately learn the perception characteristics of three types of noise at the same time. This makes the final model retain the independent perception characteristics of each type of noise while achieving accurate fitting of the overall score.

[0072] The above are merely feasible implementations of step S304 provided in this embodiment. This embodiment does not specifically limit the specific implementation of step S30.

[0073] This embodiment provides a multi-dimensional vehicle sound quality evaluation method. It acquires the historical proportion of different driving conditions for various vehicles, historical in-vehicle noise data under those driving conditions, and psychologically expected deductions for historical noise levels. Based on the historical in-vehicle noise data, it extracts historical acoustic features corresponding to the driving conditions. It obtains subjective evaluation scores from the target user group for the historical noise data and uses these scores as training labels. The initial model is trained based on the historical proportion of driving conditions, the historical acoustic features, the psychologically expected deductions for historical noise levels, and the training labels to obtain a sound quality evaluation model. This method achieves deep integration of multi-source historical data and user subjective perception, constructing an intelligent evaluation model that can accurately quantify vehicle sound quality levels in specific scenarios, providing a reliable technical means for the objective and standardized evaluation of vehicle sound quality.

[0074] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the multidimensional automotive sound quality evaluation method of this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0075] This application also provides a multi-dimensional automotive sound quality evaluation device; please refer to [reference needed]. Figure 3 The multi-dimensional vehicle sound quality evaluation device includes: The data input module 10 is used to obtain the driving condition ratio of the target vehicle, the in-vehicle noise data under the driving condition, and the psychological expectation deduction of the noise level. The feature extraction module 20 is used to extract acoustic features of multiple dimensions corresponding to road noise, wind noise and powertrain noise based on the in-vehicle noise data. The model calculation module 30 is used to input the driving condition ratio, the acoustic features of the multiple dimensions, and the psychological expectation deduction of the noise level into the sound quality evaluation model to obtain the subjective evaluation score of the sound quality of the target vehicle model.

[0076] The multi-dimensional vehicle sound quality evaluation device provided in this application, employing the multi-dimensional vehicle sound quality evaluation method described in the above embodiments, can solve the technical problem of how to achieve automated evaluation of various types of in-vehicle noise in specific driving scenarios. Compared with the prior art, the beneficial effects of the multi-dimensional vehicle sound quality evaluation device provided in this application are the same as those of the multi-dimensional vehicle sound quality evaluation method provided in the above embodiments, and other technical features in the multi-dimensional vehicle sound quality evaluation device are the same as those disclosed in the methods of the above embodiments, and will not be repeated here.

[0077] The data input module 10 is also used to acquire vehicle driving data of the target vehicle model; based on the vehicle driving data, determine the proportion of urban driving conditions, suburban driving conditions, highway driving conditions, and low-speed acceleration driving conditions of the target vehicle model; and use the proportion of urban driving conditions, suburban driving conditions, highway driving conditions, and low-speed acceleration driving conditions as the driving condition proportions.

[0078] The data input module 10 is further configured to acquire first noise data of the target vehicle when it is traveling at a first preset speed on a smooth asphalt road and a rough road, and use the first noise data as the in-vehicle noise data under road noise conditions; acquire second noise data of the target vehicle when it is traveling at a second preset speed on a smooth asphalt road, and use the second noise data as the in-vehicle noise data under wind noise conditions and powertrain noise conditions, wherein the second preset speed is higher than the first preset speed.

[0079] The data input module 10 is also used to obtain the price range of the target vehicle model; based on the in-vehicle noise data, to obtain a first actual sound pressure level and a second actual sound pressure level; based on the driving conditions corresponding to the in-vehicle noise data and the price range, to obtain a first preset standard sound pressure level and a second preset standard sound pressure level; and based on the first actual sound pressure level, the second actual sound pressure level, the first preset standard sound pressure level, and the second preset standard sound pressure level, to obtain a psychological expectation deduction for the noise level.

[0080] The feature extraction module 20 is further used to obtain the powertrain noise extraction speed interval; based on the in-vehicle noise data, extract the overall sound pressure level, overall loudness, overall speech intelligibility, overall sharpness, sound pressure level RMS value and spectrum value of different frequency bands as acoustic features corresponding to road noise; based on the in-vehicle noise data, extract the overall sound pressure level, overall loudness, overall speech intelligibility, overall sharpness and spectrum value as acoustic features corresponding to wind noise; divide the in-vehicle noise data based on the powertrain noise extraction speed interval to obtain in-vehicle noise slice data at different vehicle speeds; based on the in-vehicle noise slice data, extract the sound pressure level, sound pressure level change rate, reducer order noise prominence ratio, motor order noise prominence ratio, sound pressure level RMS value and spectrum value of preset frequency bands corresponding to different powertrain orders as acoustic features corresponding to powertrain noise.

[0081] The model calculation module 30 is further used to obtain the historical proportion of driving conditions for different vehicles, historical in-vehicle noise data under driving conditions, and psychological expectation deductions for historical noise levels; based on the historical in-vehicle noise data, extract the historical acoustic features corresponding to the driving conditions; obtain the subjective evaluation scores of the target user group on the historical noise data, and use the subjective evaluation scores as training labels; train the initial model based on the historical proportion of driving conditions, the historical acoustic features, the psychological expectation deductions for historical noise levels, and the training labels to obtain a sound quality evaluation model.

[0082] The model calculation module 30 is further configured to input the historical acoustic features into the initial model to obtain road noise score, wind noise score and total dynamic noise score; based on the road noise score, the wind noise score, the total dynamic noise score, the historical proportion of the driving condition and the psychological expectation of the historical noise level, to obtain an initial subjective evaluation score of sound quality; to compare the initial subjective evaluation score of sound quality with the training labels to obtain the model fitting rate; and to determine the sound quality evaluation model when the model fitting rate reaches a preset threshold.

[0083] This application provides a multi-dimensional vehicle sound quality evaluation device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the multi-dimensional vehicle sound quality evaluation method in the above embodiment 1.

[0084] The following is for reference. Figure 4The diagram illustrates a structural schematic of a multi-dimensional automotive sound quality evaluation device suitable for implementing embodiments of this application. The multi-dimensional automotive sound quality evaluation device in this application may include, but is not limited to, mobile terminals such as mobile phones, laptops, digital radio receivers, PDAs (Personal Digital Assistants), PADs (Portable Application Description), PMPs (Portable Media Players), in-vehicle terminals (e.g., in-vehicle navigation terminals), and fixed terminals such as digital TVs and desktop computers. Figure 4 The multi-dimensional automotive sound quality evaluation device shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of this application.

[0085] like Figure 4 As shown, the multi-dimensional automotive sound quality evaluation device may include a processing unit 1001 (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in ROM (Read Only Memory) 1002 or a program loaded from storage device 1003 into random access memory (RRAM) 1004. The RAM 1004 also stores various programs and data required for the operation of the multi-dimensional automotive sound quality evaluation device. The processing unit 1001, ROM 1002, and RAM 1004 are interconnected via a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Typically, the following systems can be connected to I / O interface 1006: input devices 1007 including, for example, touchscreens, touchpads, keyboards, mice, image sensors, microphones, accelerometers, gyroscopes, etc.; output devices 1008 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 1003 including, for example, magnetic tapes, hard disks, etc.; and communication devices 1009. Communication device 1009 allows the multi-dimensional automotive sound quality evaluation equipment to communicate wirelessly or wiredly with other devices to exchange data. Although a multi-dimensional automotive sound quality evaluation equipment with various systems is shown in the figure, it should be understood that it is not required to implement or possess all the systems shown. More or fewer systems can be implemented alternatively.

[0086] Specifically, according to the embodiments disclosed in this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments disclosed in this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device, or installed from storage device 1003, or installed from ROM 1002. When the computer program is executed by processing device 1001, it performs the functions defined in the methods of the embodiments disclosed in this application.

[0087] The multi-dimensional vehicle sound quality evaluation device provided in this application, employing the multi-dimensional vehicle sound quality evaluation method described in the above embodiments, can solve the technical problem of how to achieve automated evaluation of various types of in-vehicle noise in specific driving scenarios. Compared with the prior art, the beneficial effects of the multi-dimensional vehicle sound quality evaluation device provided in this application are the same as those of the multi-dimensional vehicle sound quality evaluation method provided in the above embodiments, and other technical features of this multi-dimensional vehicle sound quality evaluation device are the same as those disclosed in the previous embodiment method, and will not be repeated here.

[0088] It should be understood that the various parts disclosed in this application can be implemented using hardware, software, firmware, or a combination thereof. In the description of the above embodiments, specific features, structures, materials, or characteristics can be combined in any suitable manner in one or more embodiments or examples.

[0089] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0090] This application provides a computer-readable storage medium having computer-readable program instructions (i.e., a computer program) stored thereon, the computer-readable program instructions being used to execute the multi-dimensional vehicle sound quality evaluation method described in the above embodiments.

[0091] The computer-readable storage medium provided in this application may be, for example, a USB flash drive, but is not limited to, electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this embodiment, the computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, system, or device. The program code contained on the computer-readable storage medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (Radio Frequency), etc., or any suitable combination thereof.

[0092] The aforementioned computer-readable storage medium may be included in the multi-dimensional vehicle sound quality evaluation device; or it may exist independently and not be assembled into the multi-dimensional vehicle sound quality evaluation device.

[0093] The aforementioned computer-readable storage medium carries one or more programs. When the aforementioned one or more programs are executed by the multi-dimensional automotive sound quality evaluation device, the multi-dimensional automotive sound quality evaluation device: acquires the driving condition ratio of the target vehicle model, in-vehicle noise data under driving conditions, and psychological expectation deduction for noise level; based on the in-vehicle noise data, extracts acoustic features of multiple dimensions corresponding to road noise, wind noise, and powertrain noise respectively; inputs the driving condition ratio, the multiple dimensions of acoustic features, and the psychological expectation deduction for noise level into the sound quality evaluation model to obtain the subjective sound quality evaluation score of the target vehicle model.

[0094] Computer program code for performing the operations of this application can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a Local Area Network (LAN) or a Wide Area Network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0095] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0096] The modules described in the embodiments of this application can be implemented in software or hardware. The names of the modules do not necessarily limit the functionality of the unit itself.

[0097] The readable storage medium provided in this application is a computer-readable storage medium that stores computer-readable program instructions (i.e., a computer program) for executing the above-described multi-dimensional vehicle sound quality evaluation method. This solves the technical problem of how to achieve automated evaluation of various types of in-vehicle noise in specific driving scenarios. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided in this application are the same as those of the multi-dimensional vehicle sound quality evaluation method provided in the above embodiments, and will not be repeated here.

[0098] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the multidimensional vehicle sound quality evaluation method described above.

[0099] The computer program product provided in this application solves the technical problem of how to automatically evaluate various types of in-vehicle noise in specific driving scenarios. Compared with the prior art, the beneficial effects of the computer program product provided in this application are the same as those of the multi-dimensional vehicle sound quality evaluation method provided in the above embodiments, and will not be repeated here.

[0100] The above description is only a part of the embodiments of this application and does not limit the patent scope of this application. All equivalent structural transformations made under the technical concept of this application and using the contents of the specification and drawings of this application, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A multi-dimensional method for evaluating automotive sound quality, characterized in that, The method includes: Obtain the percentage of driving conditions for the target vehicle model, in-vehicle noise data under driving conditions, and psychological deductions for noise levels; Based on the in-vehicle noise data, acoustic features of multiple dimensions corresponding to road noise, wind noise and powertrain noise are extracted. The driving condition ratio, the acoustic characteristics of the multiple dimensions, and the psychological expectation deduction for noise level are input into the sound quality evaluation model to obtain the subjective sound quality evaluation score of the target vehicle model.

2. The method as described in claim 1, characterized in that, Obtain the percentage of driving conditions for the target vehicle model, including: Obtain vehicle driving data for the target model; Based on the vehicle driving data, determine the proportion of urban driving conditions, suburban driving conditions, highway driving conditions, and low-speed acceleration driving conditions for the target vehicle model. The proportion of urban driving conditions, the proportion of suburban driving conditions, the proportion of highway driving conditions, and the proportion of low-speed acceleration driving conditions are taken as the driving condition proportions.

3. The method as described in claim 1, characterized in that, The driving conditions include road noise conditions, wind noise conditions, and powertrain noise conditions; Obtain in-vehicle noise data for the target vehicle under preset driving conditions, including: Acquire the first noise data of the target vehicle when it is traveling at a constant speed of a first preset speed on a smooth asphalt road and a rough road, and use the first noise data as the in-vehicle noise data under road noise conditions; Acquire the second noise data of the target vehicle model when it is traveling at a constant speed of a second preset speed on a smooth asphalt road surface, and use the second noise data as the in-vehicle noise data for wind noise conditions and powertrain noise conditions, wherein the second preset speed is higher than the first preset speed.

4. The method as described in claim 1, characterized in that, Determine the psychological deduction for the noise level of the target vehicle model, including: Obtain the price range of the target vehicle model; Based on the in-vehicle noise data, the first actual sound pressure level and the second actual sound pressure level are obtained; Based on the driving conditions corresponding to the in-vehicle noise data and the price range, a first preset standard sound pressure level and a second preset standard sound pressure level are obtained. Based on the first actual sound pressure level, the second actual sound pressure level, the first preset standard sound pressure level, and the second preset standard sound pressure level, a psychological expectation deduction for noise level is obtained.

5. The method as described in claim 1, characterized in that, Based on the in-vehicle noise data, acoustic features corresponding to road noise, wind noise, and powertrain noise in multiple dimensions are extracted, including: Extracting vehicle speed intervals from powertrain noise; Based on the in-vehicle noise data, the total sound pressure level, total loudness, total speech intelligibility, total sharpness, sound pressure level RMS value and spectral value of different frequency bands are extracted as acoustic features corresponding to road noise. Based on the in-vehicle noise data, the total sound pressure level, total loudness, total speech intelligibility, total sharpness, and spectral values ​​are extracted as acoustic features corresponding to wind noise. Based on the powertrain noise extraction speed interval, the in-vehicle noise data is divided to obtain in-vehicle noise slice data at different vehicle speeds. Based on the in-vehicle noise slice data, the sound pressure level, sound pressure level change rate, reducer order noise prominence ratio, motor order noise prominence ratio, sound pressure level RMS value and spectrum value of the preset frequency band are extracted as acoustic features corresponding to powertrain noise for different powertrain orders.

6. The method as described in claim 1, characterized in that, Before the step of inputting the driving condition ratio, the acoustic characteristics of the multiple dimensions, and the psychological expectation deduction of the noise level into the sound quality evaluation model to obtain the subjective sound quality evaluation score of the target vehicle model, the method further includes: Obtain the historical percentage of different driving conditions for different vehicles, historical in-vehicle noise data under different driving conditions, and the expected deduction for historical noise levels. Based on the historical in-vehicle noise data, extract the historical acoustic features corresponding to the driving conditions; Obtain the subjective evaluation scores of the target user group on the historical noise data, and use the subjective evaluation scores as training labels; The initial model is trained based on the historical proportion of driving conditions, the historical acoustic characteristics, the expected deduction for historical noise levels, and the training labels to obtain a sound quality evaluation model.

7. The method as described in claim 6, characterized in that, The initial model is trained based on the historical proportion of driving conditions, the historical acoustic characteristics, the expected deduction for historical noise levels, and the training labels to obtain a sound quality evaluation model, including: The historical acoustic features are input into the initial model to obtain road noise score, wind noise score and total dynamic noise score; Based on the road noise score, the wind noise score, the total power noise score, the historical proportion of driving conditions, and the psychological expectation of historical noise levels, an initial subjective evaluation score for sound quality is obtained. The initial subjective evaluation score of sound quality is compared with the training labels to obtain the model fit rate; When the model fitting rate reaches a preset threshold, the sound quality evaluation model is determined.

8. A multi-dimensional automotive sound quality evaluation device, characterized in that, The device includes: The data input module is used to obtain the driving conditions percentage of the target vehicle model, the in-vehicle noise data under driving conditions, and the psychological expectation deduction for noise level. The feature extraction module is used to extract acoustic features of multiple dimensions corresponding to road noise, wind noise and powertrain noise based on the in-vehicle noise data. The model calculation module is used to input the driving condition ratio, the acoustic features of the multiple dimensions, and the psychological expectation deduction of the noise level into the sound quality evaluation model to obtain the subjective sound quality evaluation score of the target vehicle model.

9. A multi-dimensional automotive sound quality evaluation device, characterized in that, The device includes: a memory, a processor, and a computer program stored in the memory and executable on the processor, the computer program being configured to implement the steps of the multidimensional vehicle sound quality evaluation method as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium. When the computer program is executed by a processor, it implements the steps of the multi-dimensional vehicle sound quality evaluation method as described in any one of claims 1 to 7.

Images