A forward design method for low-noise interior of a power-centralized EMU that incorporates sound quality optimization

Abstract

This invention belongs to the field of noise control and acoustic design of rail transit vehicles, and discloses a forward design method for low-noise interior of a power-centralized EMU combined with sound quality optimization. The steps are as follows: First, determine the design target value for the in-vehicle noise sound pressure level and the target value for the comprehensive annoyance index of sound quality. Second, based on the whole-vehicle noise simulation model, allocate noise limits from the perspective of sound source-transmission path combinations, and verify the rationality of the allocation results in conjunction with engineering feasibility requirements. Third, establish independent simulation models for different sound source-transmission path combinations, determine initial optimization design parameters, and input them into the whole-vehicle noise simulation model for iterative optimization to ensure that the total A-weighted sound pressure level in the passenger compartment reaches the design target. After the sound pressure level meets the target, obtain the comprehensive annoyance index of sound quality using a sound quality evaluation and prediction model. When the sound quality does not meet the target, locate the psychoacoustic parameters affecting the sound quality and the corresponding sound source-transmission path combinations, and perform secondary optimization until the sound quality meets the target.

Description

Technical Field

[0001] This invention belongs to the field of noise control and acoustic design of rail transit vehicles, and relates to a positive design method for low noise inside a power-centralized EMU that combines sound quality optimization. Background Technology

[0002] With the rapid development of high-speed rail, centralized power EMUs have been widely used in intercity and suburban railways due to their advantages such as flexible formation, low operating costs, and convenient maintenance. However, the issue of noise inside EMUs is receiving increasing attention. Excessive noise levels not only affect passenger comfort but may also have adverse effects on the physical and mental health of train crew members.

[0003] Traditional noise control in high-speed trains relies primarily on experience-based design and subsequent modifications, lacking a systematic, forward-looking design approach. This "design first, then address" model often results in high noise control costs, limited effectiveness, and an inability to fundamentally solve the problem. Furthermore, existing technologies largely focus on controlling sound pressure level (dB(A)), neglecting sound quality, a crucial indicator reflecting passengers' subjective experience.

[0004] In recent years, scholars and engineers both domestically and internationally have conducted extensive research on noise control in rail transit vehicles. Patent application CN112163274A constructs an iterative optimization design method for in-vehicle noise, guided by optimal noise reduction costs. This method overcomes the subjective limitations of traditional experience-based design and effectively shortens the design cycle. However, this method focuses solely on the single indicator of sound pressure level, completely neglecting sound quality optimization. Furthermore, it lacks quantitative analysis of the coupling effect between sound sources and transmission paths, and the allocation of noise indicators is solely cost-oriented, lacking refinement and scientific rigor. Patent application CN107745721A establishes a comprehensive low-noise design system encompassing "target setting - indicator allocation - structural selection - simulation verification - vehicle integration," achieving pre-control of in-vehicle sound pressure level during the design phase. However, this method still uses sound pressure level compliance as the sole criterion, failing to incorporate sound quality into the entire design process. When simulation results fail to meet standards, it only iterates by simply increasing the vehicle's sound insulation, lacking multi-dimensional collaborative optimization strategies and failing to establish a dedicated secondary optimization scheme for sound quality degradation.

[0005] In summary, the existing technology has the following shortcomings: (1) It lacks a dual-closed-loop forward design method that combines sound pressure level control with sound quality optimization; (2) It does not adequately consider the coupling effect between the sound source and the transmission path, and the allocation of noise limits lacks a scientific basis; (3) The compatibility between simulation analysis methods and the "sound source-transmission path" combination needs to be improved; (4) The secondary optimization strategy when the sound quality does not meet the standard is not clear enough, making it difficult to quickly locate the core problem and formulate targeted measures. Therefore, it is urgent to propose a systematic, scientific, and operable forward design method for low noise inside the power-centralized EMU to achieve dual compliance of sound pressure level and sound quality. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides a forward design method for low-noise interior design of power-centralized EMU trains that incorporates sound quality optimization.

[0007] The technical solution of this invention: A forward design method for low-noise interior of a power-centralized EMU that incorporates sound quality optimization includes the following steps: S1. Determine the design target value of the sound pressure level and the target value of the comprehensive annoyance index of the sound quality inside the power-centralized EMU. S1.1 Select no fewer than three power-centralized EMUs of the same type, speed class and good operating condition as sample vehicles, and conduct whole-vehicle line tests on line sections including straight sections, curved sections, bridge sections and tunnel sections. S1.2 Acoustic measurement points were set up inside the passenger compartment of each carriage of the sample vehicle. A total of 6 acoustic measurement points were set up in each carriage, arranged longitudinally along the carriage. The longitudinal positions of the measurement points were set at distances of 0, 1 / 8L, 1 / 4L, 1 / 2L, 3 / 4L, and 7 / 8L from the end of the carriage, respectively, where L is the total longitudinal length of the carriage. Each acoustic measurement point was located on the transverse centerline of the carriage and vertically at a height of 1.6m above the floor to simulate the actual height of the human ear above the ground. The original noise audio signals were simultaneously collected at each measurement point. The signal duration is no less than 30 seconds. After removing transient interference segments at the beginning and end of the original audio signal, a 5-second steady-state segment in the middle is uniformly extracted as the analysis sample. Subsequently, the signal is processed by acoustic software, and invalid data is removed to obtain objective parameter data for each analysis sample. The objective parameter data includes A-weighted sound pressure level and psychoacoustic parameters, including loudness, sharpness, roughness, tone modulation, and fluctuation, to complete the objective test data collection of noise in the passenger compartment of each carriage of the sample vehicle. Then, the A-weighted sound pressure level of each analysis sample is averaged to obtain the measured and verified value of the total A-weighted sound pressure level in the passenger compartment. S1.3 Conduct a subjective evaluation test: Organize no fewer than 30 reviewers aged 18-50 years old, with a male-to-female ratio of 1:1, and with normal hearing; use a rating scale to evaluate the subjective annoyance level of the analyzed samples, with specific scoring details shown in Table 1; after removing invalid data, obtain the subjective evaluation test results for each analyzed sample, and take the arithmetic mean of the subjective evaluation test results as the benchmark value for the subjective annoyance level of in-vehicle noise of the existing vehicle models. Table 1. Details of Subjective Annoyance Level Classification

[0008] S1.4. Based on the vehicle parameters, design speed, formation, car body materials, and sound insulation requirements of the power-centralized EMU, and in conjunction with the measured and verified value of the total A-weighted sound pressure level in the passenger compartment, determine the design target value for the in-car noise sound pressure level. The design target value for the in-car noise sound pressure level is 3-5 dB(A) lower than the measured and verified value of the total A-weighted sound pressure level in the passenger compartment, and is measured in dB(A). At the same time, considering the passengers' requirements for the subjective comfort of the power-centralized EMU, and with reference to the subjective annoyance benchmark value of the in-car noise of the existing vehicle model, determine the target value for the comprehensive annoyance index of sound quality. The target value for the comprehensive annoyance index of sound quality is 1-2 points lower than the subjective annoyance benchmark value of the in-car noise of the existing vehicle model.

[0009] S2. Based on the design target value of the in-vehicle noise sound pressure level, decompose the total noise value from the perspective of the combination of sound source and transmission path, allocate the design target value of the in-vehicle noise sound pressure level to the noise limit value corresponding to each sound source-transmission path combination, and verify the rationality of the noise limit allocation result. S2.1 Establish a vehicle noise simulation model based on statistical energy analysis; The vehicle noise simulation model includes a structural subsystem, an acoustic cavity subsystem, and a sound source subsystem. The structural subsystem consists of the main structural components that make up the body of the power-centralized EMU, the acoustic cavity subsystem consists of the acoustic cavities inside and under the power-centralized EMU, and the sound source subsystem consists of the noise sources that generate noise excitation during the operation of the power-centralized EMU. When dividing the vehicle noise simulation model into subsystems, the power-centralized EMU body structure is divided into 6 structural subsystems: floor, roof, side walls, end walls, window area, and door area; the acoustic cavity is divided into 4 acoustic cavity subsystems: passenger compartment acoustic cavity, driver's cab acoustic cavity, equipment compartment acoustic cavity, and undercarriage space acoustic cavity; and the noise sources are divided into 4 noise source subsystems: wheel-rail noise source, traction equipment noise source, auxiliary equipment noise source, and aerodynamic noise source; among them, the auxiliary equipment noise source is the air conditioning system noise source. The noise transmission relationships between each structural subsystem, each acoustic cavity subsystem, and each sound source subsystem are established. Among them, the vibration generated by the sound source subsystem is transmitted to the passenger compartment through the structural subsystem to characterize the structural sound transmission path; the noise generated by the sound source subsystem is transmitted to the passenger compartment through the acoustic cavity subsystem to characterize the airborne sound transmission path. This characterizes the process of vibration and noise propagating to the passenger compartment through the structural sound transmission path and the airborne sound transmission path, respectively, and thus constructs a vehicle noise simulation model. The input parameters of the whole vehicle noise simulation model were all obtained through experimental testing; among them, the material loss factor was obtained through the damping test method, the sound absorption coefficient was obtained through the standing wave tube method, the structural dynamic characteristic parameters were obtained through modal testing, and the sound source radiated sound power was obtained through the sound intensity method in a semi-anechoic chamber. The above input parameters are imported into the vehicle noise simulation model to obtain the simulation result of the total A-weighted sound pressure level in the passenger compartment. When the deviation between the simulation result of the total A-weighted sound pressure level in the passenger compartment and the measured verification value of the total A-weighted sound pressure level in the passenger compartment obtained in step S1.2 is ≤5dB(A), the vehicle noise simulation model is deemed valid. When the deviation between the simulation result of the total A-weighted sound pressure level in the passenger compartment and the measured verification value of the total A-weighted sound pressure level in the passenger compartment obtained in step S1.2 is >5dB(A), the test is repeated to correct the input parameters of the vehicle noise simulation model until the deviation between the simulation result of the total A-weighted sound pressure level in the passenger compartment and the measured verification value of the total A-weighted sound pressure level in the passenger compartment is ≤5dB(A). Subsequently, based on the valid vehicle noise simulation model, the noise limit allocation and rationality verification of each sound source-transmission path combination are performed. S2.2. Based on an effective vehicle noise simulation model, obtain the A-weighted sound pressure level contribution value of each sound source-transmission path combination in the total noise value, and determine the initial weight allocation of each sound source-transmission path combination based on the A-weighted sound pressure level contribution value: The sound source-transmission path combination includes at least the following 6 paths, covering more than 95% of the energy contribution sources of in-vehicle noise, specifically: 1) Wheel-rail noise - structural transmission path; 2) Wheel-rail noise - air transmission path; 3) Traction equipment noise - structural transmission path; 4) Traction equipment noise - air transmission path; 5) Aerodynamic noise - air transmission path; 6) Air conditioning system noise - air transmission path. The initial weighting formula for each sound source-transmission path combination is as follows: in, The initial weights are assigned to the j-th sound source-transmission path combination, where n is the total number of sound source-transmission path combinations. The A-weighted sound pressure level contribution value for the j-th sound source-transmission path combination; S2.3. Based on the initial weight allocation obtained in step S2.2, and considering the coupling correction coefficient... The noise limit for each sound source-transmission path combination is derived by inversely using the principle of energy superposition. ; The formula for superposition of energy is as follows: in, The target value for the sound pressure level of in-vehicle noise; This is the noise limit for the i-th sound source-transmission path combination; It is the coupling correction coefficient for the i-th sound source-transmission path combination, used to correct the nonlinear effect of energy superposition caused by multi-sound source coupling; The larger the value, the greater the actual contribution of the sound source-transmission path combination to the total noise value under the combined action of multiple sound sources. The value ranges from 0.85 to 1.15; Noise limits for each sound source-transmission path combination Based on the initial weight allocation Determine the initial weight allocation. The larger the value, the greater the contribution of the sound source-transmission path combination to the total A-weighted sound pressure level in the guest room; S2.4 Noise limits for each source-transmission path combination obtained by reverse calculation Conduct a rationality verification; set noise limits for each sound source-transmission path combination. The noise limits of each sound source-transmission path combination are compared with the achievable lower noise limit for the corresponding sound source-transmission path combination to determine whether the noise limit meets the engineering feasibility requirements. The achievable lower noise limit refers to the lowest noise value that the corresponding sound source-transmission path combination can actually achieve after noise reduction design, under the existing achievable low-noise design level of similar power-centralized EMUs. If the noise limits for any sound source-transmission path combination are not lower than the achievable lower noise limit for that combination, then the noise limit allocation scheme is considered reasonable; if the noise limit for any sound source-transmission path combination is lower than the achievable lower noise limit for that combination, then the noise limit allocation scheme is considered reasonable. If all noise levels are below the achievable lower limit of the corresponding sound source-transmission path combination, the noise limit of that sound source-transmission path combination shall be adjusted to its achievable lower limit. The remaining noise reduction shall be allocated to other sound source-transmission path combinations until the noise limits of each sound source-transmission path combination meet the feasibility requirements of the project.

[0010] S3. To address the differences in noise characteristics among various sound source-transmission path combinations, independent simulation models are established, each focusing on a single sound source-transmission path combination and using A-weighted sound pressure level as the output. Considering that transmission path types include both airborne and structural transmission paths, and the noise frequency range covers 20Hz to 20000Hz, a suitable simulation analysis method is first selected based on the transmission path type and frequency range. Then, corresponding independent simulation models are established based on the sound source location, noise propagation path, and the receiving location in the living room. Subsequently, key design parameters in each sound source-transmission path combination are optimized to meet the corresponding noise limits, thus determining the initial optimized design parameters for each sound source-transmission path combination. S3.1. Select appropriate simulation analysis methods based on the transmission path type and frequency range of each sound source-transmission path combination. Specifically, for sound source-transmission path combinations that transmit sound through the air, the acoustic boundary element method (BEM) is used for simulation analysis in the 20Hz~1000Hz frequency band, and the statistical energy analysis (SEA) method is used for simulation analysis in the 1000Hz~20000Hz frequency band. For sound source-transmission path combinations that transmit sound through structures, the finite element method (FEM) is used for simulation analysis in the 20Hz~500Hz frequency band, the transfer path analysis (TPA) method is used for simulation analysis in the 500Hz~2000Hz frequency band, and the statistical energy analysis method is used for simulation analysis in the 2000Hz~20000Hz frequency band. S3.2. Based on the simulation analysis method determined in step S3.1, establish independent simulation models for each sound source-transmission path combination. Specifically, for sound source-transmission path combinations transmitted through the air, establish an airborne sound transmission simulation model to simulate the process of sound source noise being transmitted through the air to the passenger compartment. The input parameters of the airborne sound transmission simulation model include at least the sound source power, vehicle body sound insulation, and passenger compartment sound absorption coefficient. For sound source-transmission path combinations transmitted through the structure, establish a structural sound transmission simulation model to simulate the process of sound source vibration excitation being transmitted through the vehicle body structure and forming noise in the passenger compartment. The input parameters of the structural sound transmission simulation model include at least the sound source vibration excitation parameters, vehicle body structural material parameters, and structural damping parameters. The vehicle body structural material parameters include elastic modulus, Poisson's ratio, and density. S3.3 Run the independent simulation model of each sound source-transmission path combination to obtain the A-weighted sound pressure level contribution value of the corresponding sound source-transmission path combination; compare the obtained A-weighted sound pressure level contribution value with the noise limit value corresponding to each sound source-transmission path combination obtained in step S2, and adjust and optimize the key design parameters in each sound source-transmission path combination to obtain the initial optimized design parameters corresponding to each sound source-transmission path combination, subject to the constraint of meeting the corresponding noise limit value.

[0011] S4. Input the initial optimization design parameters corresponding to each sound source-transmission path combination into the vehicle noise simulation model, perform vehicle noise simulation analysis, and obtain the vehicle noise simulation results. Based on the vehicle noise simulation results, further optimize the initial optimization design parameters of each sound source-transmission path combination under the vehicle integration condition until the vehicle noise meets the design target value of the in-vehicle noise sound pressure level. S4.1 Input the initial optimization design parameters corresponding to each sound source-transmission path combination into the vehicle noise simulation model, run the vehicle noise simulation model, and output the total A-weighted sound pressure level in the passenger compartment; compare the total A-weighted sound pressure level in the passenger compartment with the design target value of the in-vehicle noise sound pressure level. When the deviation between the two is ≤0.5dB(A), the sound pressure level is determined to meet the standard, and proceed to step S5; when the deviation between the two is >0.5dB(A), start the iterative optimization process based on parameter sensitivity ranking. S4.2 During iterative optimization, parameters are adjusted in descending order of sensitivity. First, the initial optimization design parameters for each sound source-transmission path combination are changed, and the corresponding changes in vehicle noise are calculated to obtain the sensitivity coefficient of each initial optimization design parameter to vehicle noise. The sensitivity coefficient characterizes the degree of influence of changes in the initial optimization design parameters on vehicle noise. Initial optimization design parameters with larger sensitivity coefficients are adjusted first to make the vehicle noise simulation results approximate the design target value of the in-vehicle noise sound pressure level. Subsequently, the initial optimization design parameters with smaller sensitivity coefficients are further adjusted to improve the accuracy of parameter adjustment. S4.3 After each round of initial optimization design parameter adjustment, the vehicle noise simulation model is run again to check the total A-weighted sound pressure level in the passenger compartment; the sound pressure level is re-determined to meet the standard; if the requirements are still not met after 5 consecutive iterations, return to step S2 and redistribute the noise limit values ​​for each sound source-transmission path combination.

[0012] S5. Extract the vehicle noise after the sound pressure level meets the standard, and extract the psychoacoustic parameters of the vehicle noise through acoustic analysis software; then, based on the objective parameter data of each analysis sample obtained in step S1.2 and the subjective evaluation test results of each analysis sample obtained in step S1.3, establish a multiple linear regression sound quality evaluation prediction model, and input the psychoacoustic parameters of the vehicle noise after the sound pressure level meets the standard into the multiple linear regression sound quality evaluation prediction model to obtain the corresponding comprehensive annoyance index of sound quality, which is used as the sound quality evaluation result. The multiple linear regression sound quality evaluation and prediction model is expressed as follows: in, The overall annoyance index for sound quality ranges from 1 to 10 points. For loudness, For sharpness, For roughness, For sound scheduling, For volatility; , , , , , The regression coefficients are used to fit and calibrate the subjective evaluation test results using the least squares method.

[0013] S6. Compare the overall annoyance index of sound quality obtained in step S5 with the target value of the overall annoyance index of sound quality. When the overall annoyance index of sound quality does not reach the target value of the overall annoyance index of sound quality, identify the psychoacoustic parameter that has the greatest impact on the overall annoyance index of sound quality through correlation analysis. Then, locate the sound source-transmission path combination corresponding to the psychoacoustic parameter based on the energy contribution method of transmission path analysis. Subsequently, with the goal of reducing the adverse effect of the psychoacoustic parameter on the overall annoyance index of sound quality, perform secondary optimization on the initial optimization design parameters of the sound source-transmission path combination, and repeat steps S4 to S5. Then return to step S6 until the overall annoyance index of sound quality reaches the target value of the overall annoyance index of sound quality, thus completing the double closed-loop forward design process. S6.1. Compare the overall annoyance index of sound quality obtained in step S5 with the target value of the overall annoyance index of sound quality. When the overall annoyance index of sound quality reaches the target value, the design process ends. When the overall annoyance index of sound quality does not reach the target value, proceed to secondary optimization. By calculating the correlation coefficient between the psychoacoustic parameters of each analysis sample obtained in step S1.2 and the overall annoyance index M of sound quality, identify the psychoacoustic parameter that has the greatest impact on the overall annoyance index of sound quality. Then, for this psychoacoustic parameter, use the energy contribution method based on transmission path analysis to calculate the contribution ratio of each sound source-transmission path combination to this psychoacoustic parameter, and take the sound source-transmission path combination with the highest contribution ratio as the object of secondary optimization. With the goal of reducing the adverse effect of this psychoacoustic parameter on the overall annoyance index of sound quality, adjust the design parameters of the object of secondary optimization to obtain the design parameters after secondary optimization. S6.2 Update the optimized design parameters to the vehicle noise simulation model, repeat steps S4 to S5 to obtain a new comprehensive noise level index, and then return to step S6.1. When the comprehensive noise level index reaches the target value, the design process ends; when the comprehensive noise level index still does not reach the target value, repeat the secondary optimization process until the comprehensive noise level index reaches the target value.

[0014] The beneficial effects of this invention are as follows: This invention constructs a forward design method for low-noise in-vehicle design of EMUs that combines sound pressure level closed-loop and sound quality closed-loop, which can simultaneously consider the in-vehicle noise sound pressure level target and passengers' subjective sound quality experience during the design stage, reducing later rectification and repeated verification; This invention allocates noise limits from the perspective of sound source-transmission path combination and combines it with engineering feasibility verification, making the noise control target more reasonable and feasible; This invention selects appropriate simulation analysis methods according to the transmission path type and frequency range, improving the pertinence of simulation analysis and parameter optimization; This invention evaluates the optimization results through the comprehensive annoyance index of sound quality and carries out secondary optimization when the sound quality does not meet the standard, which is conducive to achieving dual compliance of in-vehicle noise sound pressure level and sound quality. Attached Figure Description

[0015] Figure 1 This is a flowchart of a low-noise forward design method for a power-centralized EMU that incorporates sound quality optimization, as disclosed in an embodiment of the present invention. Figure 2 This is a flowchart of the noise limit allocation and rationality verification of the sound source-transmission path combination disclosed in the embodiments of the present invention; Figure 3The flowchart below shows the establishment of a sound quality evaluation and prediction model based on subjective and objective data as disclosed in the embodiments of the present invention. Detailed Implementation

[0016] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings and technical solutions.

[0017] like Figure 1 As shown in the figure, this embodiment exemplarily demonstrates a low-noise forward design method for power-centralized EMU trains that incorporates sound quality optimization, including the following steps: 1) Collection of subjective and objective benchmark data for in-vehicle noise: First, select no fewer than three power-centralized EMUs of the same type and speed level, and in good operating condition, as sample vehicles. Before testing, check the operation status of the doors, windows, air conditioning system, traction equipment, and undercarriage equipment of the sample vehicles to ensure that the vehicles are in normal operating condition. The test route is selected from typical sections of the line, including straight sections, curved sections, bridge sections, and tunnel sections. During the test, try to choose a time period with stable weather, low wind speed, and no obvious sudden external noise interference to reduce the impact of environmental factors on the test results.

[0018] Six acoustic measurement points were installed inside the passenger compartment of each carriage. These points were arranged longitudinally along the carriage, with their longitudinal positions starting from the end of the carriage and set at distances of 0, 1 / 8L, 1 / 4L, 1 / 2L, 3 / 4L, and 7 / 8L from that starting point, where L is the total longitudinal length of the carriage. Laterally, each measurement point was located on the transverse centerline of the carriage, and vertically, it was positioned 1.6m above the floor to simulate the typical sound field location near a passenger's ear. After the acoustic measurement points were installed, the acoustic acquisition equipment was calibrated, and the measurement point number, carriage number, and corresponding acquisition channel were recorded.

[0019] Raw noise audio signals were simultaneously acquired at each acoustic measurement point, with each segment lasting at least 30 seconds. During acquisition, the vehicle's operating section, speed, and test time were simultaneously marked using vehicle operation records or manual recording to ensure that subsequent noise samples corresponded to the vehicle's operating conditions. After acquisition, each channel signal was checked for saturation, breaks, sudden impacts, or significant background interference; any abnormal signal segments were discarded. For valid signal segments, transient interference segments at the beginning and end of the raw noise audio signal were removed, and a uniform 5-second steady-state segment was extracted as the analysis sample. Subsequently, acoustic analysis software was used to extract the A-weighted sound pressure level and psychoacoustic parameters for each analysis sample. Psychoacoustic parameters included loudness, sharpness, roughness, tone modulation, and fluctuation. Finally, the A-weighted sound pressure levels of each valid analysis sample were averaged to obtain the measured and verified value of the total A-weighted sound pressure level in the passenger compartment.

[0020] In the subjective evaluation test, no fewer than 30 hearing-normal reviewers were selected. Before the evaluation, the reviewers were given an explanation of the scoring rules and familiarized with the subjective annoyance rating scale of 1 to 10 points. The specific scoring details are shown in Table 1.

[0021] Table 1. Details of Subjective Annoyance Level Classification

[0022] Meanwhile, the output sound pressure level of the professional monitoring headphones used in the subsequent subjective evaluation tests was calibrated, and the playback sound pressure level should be calibrated to be the same as the sound pressure level at the original acquisition point in the living room. All samples used the same playback equipment and parameter settings, and the headphone volume was not allowed to be adjusted during the evaluation process.

[0023] Subjective evaluation tests were conducted in a standard listening environment with background noise below 30 dB(A). Reviewers played the analysis samples through calibrated professional monitoring headphones. After each analysis sample was played, the reviewers scored it based on their subjective feelings. Data with obviously missing scores, duplicate abnormal scores, or scores that were significantly inconsistent with the overall evaluation trend were removed. Valid scores were retained and their arithmetic mean was calculated as the benchmark value for subjective annoyance of in-vehicle noise in existing vehicle models.

[0024] 2) Determination of Design Target Values: Regarding the sound pressure level target value, based on the vehicle parameters, design speed, formation, car body materials, and sound insulation configuration requirements of the power-centralized EMU, and combined with the measured verification value of the total A-weighted sound pressure level in the passenger compartment, the design target value for the in-car noise sound pressure level is determined. The design target value for the in-car noise sound pressure level is 3-5 dB(A) lower than the measured verification value of the total A-weighted sound pressure level in the passenger compartment, and is measured in dB(A). Regarding the sound quality target value, considering passengers' requirements for the subjective comfort of the power-centralized EMU, and referencing the subjective annoyance benchmark value of in-car noise in existing models, the target value for the comprehensive annoyance index of sound quality is determined. The target value for the comprehensive annoyance index of sound quality is 1-2 points lower than the subjective annoyance benchmark value of in-car noise in existing models.

[0025] 3) Establishment and Verification of the Vehicle Noise Simulation Model: A vehicle noise simulation model was established based on the statistical energy analysis method. Before modeling, basic data such as the body structure dimensions, material parameters, carriage layout, door and window positions, equipment compartment layout, and the locations of major noise sources of the target power-centralized EMU were collected. Based on the above basic data, a vehicle noise simulation model was established in the acoustic simulation software, and receiving positions corresponding to the measured acoustic measurement points were set in the passenger compartment acoustic cavity.

[0026] When dividing the model, the body structure of the power-centralized EMU is divided into structural subsystems such as floor, roof, side walls, end walls, window area and door area; the acoustic cavity is divided into passenger compartment acoustic cavity, driver's cab acoustic cavity, equipment compartment acoustic cavity and undercarriage space acoustic cavity; the noise source is divided into wheel-rail noise source, traction equipment noise source, auxiliary equipment noise source and aerodynamic noise source, among which the auxiliary equipment noise source is the air conditioning system noise source.

[0027] During modeling, based on the car body structure layout, equipment installation locations, and passenger compartment space locations of the target power-centralized EMU, the correspondence between the sound source subsystem, structural subsystem, and acoustic cavity subsystem is established in the whole-vehicle noise simulation model. For noise sources that transmit to the car body in the form of vibration, their location of action is set on the adjacent floor, side wall, end wall, or equipment installation area structural subsystem, so that the vibration is transmitted along the car body structure towards the passenger compartment, which is used to characterize the structural sound transmission path. For noise sources that transmit to the passenger compartment in the form of airborne noise, their location of action is set in the acoustic cavity under the car, the acoustic cavity of the equipment compartment, the acoustic cavity of the driver's cab, or the acoustic cavity of the passenger compartment, so that the airborne noise is transmitted to the passenger compartment through the corresponding acoustic cavity, which is used to characterize the airborne sound transmission path.

[0028] The input parameters for the vehicle noise simulation model are obtained through experimental testing. During implementation, damping tests are performed on vehicle body panels or typical structural components to obtain material loss factors; standing wave tube tests are performed on passenger interior trim materials, sound-absorbing materials, or equivalent sound-absorbing components to obtain sound absorption coefficients; modal tests are performed on vehicle body structures or typical connecting components to obtain structural dynamic characteristic parameters; and sound intensity tests or equivalent sound source parameter tests are performed on wheel-rail noise sources, traction equipment noise sources, air conditioning system noise sources, and aerodynamic noise sources to obtain sound source radiated sound power. After the tests are completed, each input parameter is entered into the vehicle noise simulation model according to the corresponding structural subsystem, acoustic cavity subsystem, and sound source subsystem.

[0029] After completing the vehicle noise simulation model, run the model to obtain the simulated total A-weighted sound pressure level (SPL) in the passenger compartment. Compare this simulation result with the measured APL of the passenger compartment. If the deviation is ≤5 dB(A), the vehicle noise simulation model is deemed to meet the engineering prediction accuracy requirements and can be used for subsequent noise limit allocation. If the deviation is >5 dB(A), check the source radiated sound power, absorption coefficient, material loss factor, and structural dynamic characteristic parameters. Based on the check results, retest or correct the model input parameters until the deviation between the simulation result and the measured value is ≤5 dB(A).

[0030] 4) Noise limit allocation and rationality verification for each sound source-transmission path combination: Based on the vehicle noise simulation model, the A-weighted sound pressure level contribution value of each sound source-transmission path combination in the total noise value is extracted. The larger the contribution value, the greater the impact of the combination on the total A-weighted sound pressure level of the passenger compartment.

[0031] The sound source-transmission path combination includes at least the following six paths: wheel-rail noise-structure transmission path, wheel-rail noise-air transmission path, traction equipment noise-structure transmission path, traction equipment noise-air transmission path, aerodynamic noise-air transmission path, and air conditioning system noise-air transmission path. These paths cover the main energy contribution sources of in-vehicle noise.

[0032] The initial allocation weights are calculated based on the contribution values ​​of each combination and used as the basis for noise limit allocation.

[0033] The initial weighting formula for each sound source-transmission path combination is as follows: in, The initial weights are assigned to the j-th sound source-transmission path combination, where n is the total number of sound source-transmission path combinations. The A-weighted sound pressure level contribution value for the j-th sound source-transmission path combination; Then, weights are initially assigned to each combination, while considering the coupling correction coefficient. The noise limit for each sound source-transmission path combination is derived by inversely using the principle of energy superposition. ; The formula for superposition of energy is as follows: in, The target value for the sound pressure level of in-vehicle noise; This is the noise limit for the i-th sound source-transmission path combination; It is the coupling correction coefficient for the i-th sound source-transmission path combination, used to correct the nonlinear effect of energy superposition caused by multi-sound source coupling; The larger the value, the greater the actual contribution of the sound source-transmission path combination to the total noise value under the combined action of multiple sound sources. The value ranges from 0.85 to 1.15; Noise limits for each sound source-transmission path combination Based on the initial weight allocation Determine the initial weight allocation. The larger the value, the greater the contribution of the sound source-transmission path combination to the total A-weighted sound pressure level in the guest room; After the noise limit is derived, the noise limit for each sound source-transmission path combination is compared with the corresponding engineering-achievable lower noise limit. The engineering-achievable lower noise limit refers to the lowest noise level that the combination can actually achieve after noise reduction design, under the existing achievable low-noise design level of similar power-centralized EMUs. If the noise limit of each combination is not lower than the corresponding engineering-achievable lower noise limit, the noise limit allocation scheme is deemed reasonable. If the noise limit of a combination is lower than its engineering-achievable lower noise limit, it indicates that the control target assigned to that combination is too strict. The noise limit of that combination needs to be adjusted to its engineering-achievable lower noise limit, and the remaining noise reduction after adjustment is allocated to other sound source-transmission path combinations until the noise limit of each combination meets the engineering feasibility requirements.

[0034] 5) Establishment of independent simulation models for each combination and determination of initial optimization design parameters: After completing the noise limit allocation for each sound source-transmission path combination, an independent simulation model is established for each single sound source-transmission path combination. During modeling, the sound source location, transmission path, and receiving location in the guest room corresponding to the combination are retained, and the A-weighted sound pressure level corresponding to the combination is used as the model output to determine whether the combination meets the corresponding noise limit.

[0035] Based on the different transmission path types and frequency ranges, appropriate simulation analysis methods are selected for each sound source-transmission path combination. For sound source-transmission path combinations transmitted through the air, the acoustic boundary element method is used in the 20Hz~1000Hz frequency band, and the statistical energy analysis method is used in the 1000Hz~20000Hz frequency band. For sound source-transmission path combinations transmitted through structures, the finite element method is used in the 20Hz~500Hz frequency band, the transmission path analysis method is used in the 500Hz~2000Hz frequency band, and the statistical energy analysis method is used in the 2000Hz~20000Hz frequency band.

[0036] For the airborne sound transmission path, an airborne sound transmission simulation model is established to simulate the process of sound source noise being transmitted through the air to the passenger compartment. During implementation, the sound source power is used as the sound source input, the vehicle body sound insulation is used as the attenuation parameter during the process of airborne noise entering the passenger compartment, the passenger compartment sound absorption coefficient is used as the sound energy attenuation parameter inside the passenger compartment, and the A-weighted sound pressure level contribution value of this combination is output at the receiving position in the passenger compartment.

[0037] For the structural sound transmission path, a structural sound transmission simulation model is established to simulate the process of sound source vibration excitation being transmitted through the vehicle body structure and forming noise in the passenger compartment. During implementation, the sound source vibration excitation parameters are used as vibration inputs, the vehicle body structural material parameters are used to describe the structural transmission characteristics, and the structural damping parameters are used to describe the vibration attenuation characteristics. The combined A-weighted sound pressure level contribution value is output at the receiving position in the passenger compartment.

[0038] After running each independent simulation model, the A-weighted sound pressure level contribution value of the corresponding sound source-transmission path combination is obtained. This contribution value is compared with the noise limit corresponding to the combination. If the contribution value is higher than the noise limit, the relevant design parameters of the combination are adjusted; if the contribution value meets the noise limit, the current parameters are retained. After the adjustment is completed, the initial optimized design parameters corresponding to each sound source-transmission path combination are obtained.

[0039] 6) Vehicle Noise Simulation Analysis and Sound Pressure Level Iterative Optimization: The initial optimized design parameters corresponding to each sound source-transmission path combination are input into the vehicle noise simulation model to perform vehicle noise simulation analysis and output the total A-weighted sound pressure level in the passenger compartment. The total A-weighted sound pressure level in the passenger compartment is compared with the design target value of the in-vehicle noise sound pressure level. When the deviation between the two is ≤0.5dB(A), the sound pressure level is judged to meet the standard, and the sound quality evaluation stage is entered; when the deviation is >0.5dB(A), iterative optimization is initiated.

[0040] During iterative optimization, parameters are adjusted in descending order of sensitivity. Specifically, the initial optimization design parameters for each sound source-transmission path combination are changed, and the change in the vehicle noise simulation results after each parameter change is recorded. This yields the sensitivity coefficient of each parameter to the vehicle noise. A larger sensitivity coefficient indicates a more significant impact of the parameter change on the vehicle noise simulation results.

[0041] In each round of optimization, the initial optimization design parameters with larger sensitivity coefficients are adjusted first to bring the vehicle noise simulation results as close as possible to the design target value for the in-vehicle noise sound pressure level. Then, the initial optimization design parameters with smaller sensitivity coefficients are adjusted to refine and correct the simulation results. After each round of parameter adjustment, the vehicle noise simulation model is rerun, and the total A-weighted sound pressure level in the passenger compartment, the deviation from the design target value, and the parameters adjusted in this round are recorded. If the sound pressure level target is still not reached after 5 consecutive iterations, the process returns to the sound source-transmission path combined noise limit allocation step to readjust the noise limits for each combination.

[0042] 7) Establishment of a multiple linear regression sound quality evaluation and prediction model and secondary optimization of sound quality: Extract the vehicle noise after the sound pressure level meets the standard, and extract its psychoacoustic parameters using acoustic analysis software. The psychoacoustic parameters include loudness, sharpness, roughness, tone tuning, and fluctuation.

[0043] Based on the psychoacoustic parameters of each obtained analysis sample as independent variables and the corresponding subjective evaluation test results as dependent variables, a multiple linear regression sound quality evaluation prediction model is established.

[0044] The multiple linear regression sound quality evaluation and prediction model is expressed as follows: in, The overall annoyance index for sound quality ranges from 1 to 10 points. For loudness, For sharpness, For roughness, For sound scheduling, For volatility; , , , , , The regression coefficients are used to fit and calibrate the subjective evaluation test results using the least squares method.

[0045] After the model fitting is completed, check the degree of fit between the model predictions and the subjective evaluation test results to ensure that the model can be used for subsequent prediction of the comprehensive annoyance index of sound quality.

[0046] After the sound pressure level meets the standard, the psychoacoustic parameters of the vehicle noise are input into a multiple linear regression sound quality evaluation and prediction model to obtain the corresponding comprehensive sound quality annoyance index. This index is then compared with the target value of the comprehensive sound quality annoyance index. If the comprehensive sound quality annoyance index reaches the target value, the sound quality is deemed to meet the standard, and the design process ends. If the comprehensive sound quality annoyance index does not reach the target value, a secondary optimization process is initiated.

[0047] In the secondary optimization, the correlation coefficients between each psychoacoustic parameter and the overall annoyance index of sound quality are first calculated to identify the psychoacoustic parameter with the greatest impact on the overall annoyance index. Then, the energy contribution method based on transmission path analysis is used to calculate the contribution ratio of each sound source-transmission path combination to this psychoacoustic parameter, and the sound source-transmission path combination with the highest contribution ratio is selected as the object of secondary optimization, adjusting its design parameters. Subsequently, the optimized design parameters are updated in the vehicle noise simulation model, and the vehicle noise simulation and sound quality evaluation are re-executed to obtain a new overall annoyance index of sound quality.

[0048] If the new overall sound quality annoyance index reaches the target value, the design process ends; if it still does not reach the target value, a second optimization is performed until the overall sound quality annoyance index reaches the target value.

[0049] Figure 2 This is a flowchart of the noise limit allocation and rationality verification of the sound source-transmission path combination disclosed in the embodiments of the present invention; Figure 3 The flowchart below shows the establishment of a sound quality evaluation and prediction model based on subjective and objective data as disclosed in the embodiments of the present invention.

[0050] The technical solutions of the present invention are not limited to the specific embodiments described above. Any technical modifications made in accordance with the technical solutions of the present invention fall within the protection scope of the present invention.

Claims

1. A method for low-noise forward design of power-centralized EMU train interiors combining sound quality optimization, characterized in that, Includes the following steps: S1. Determine the design target value of the sound pressure level and the target value of the comprehensive annoyance index of the sound quality inside the power-centralized EMU. S2. Based on the design target value of the in-vehicle noise sound pressure level, decompose the total noise value from the perspective of the combination of sound source and transmission path, allocate the design target value of the in-vehicle noise sound pressure level to the noise limit value corresponding to each sound source-transmission path combination, and verify the rationality of the noise limit allocation result. S3. To address the differences in noise characteristics among various sound source-transmission path combinations, independent simulation models are established, each focusing on a single sound source-transmission path combination and using A-weighted sound pressure level as the output. Considering that transmission path types include both airborne and structural transmission paths, and the noise frequency range covers 20Hz to 20000Hz, a suitable simulation analysis method is first selected based on the transmission path type and frequency range. Then, corresponding independent simulation models are established based on the sound source location, noise propagation path, and the receiving location in the living room. Subsequently, key design parameters in each sound source-transmission path combination are optimized to meet the corresponding noise limits, thus determining the initial optimized design parameters for each sound source-transmission path combination. S4. Input the initial optimization design parameters corresponding to each sound source-transmission path combination into the vehicle noise simulation model, perform vehicle noise simulation analysis, and obtain the vehicle noise simulation results. Based on the vehicle noise simulation results, further optimize the initial optimization design parameters of each sound source-transmission path combination under the vehicle integration condition until the vehicle noise meets the design target value of the in-vehicle noise sound pressure level. S5. Extract the vehicle noise after the sound pressure level meets the standard, and extract the psychoacoustic parameters of the vehicle noise through acoustic analysis software; then, based on the objective parameter data of each analysis sample obtained in step S1.2 and the subjective evaluation test results of each analysis sample obtained in step S1.3, establish a multiple linear regression sound quality evaluation prediction model, and input the psychoacoustic parameters of the vehicle noise after the sound pressure level meets the standard into the multiple linear regression sound quality evaluation prediction model to obtain the corresponding comprehensive annoyance index of sound quality, which is used as the sound quality evaluation result. S6. Compare the overall sound quality annoyance index obtained in step S5 with the target value of the overall sound quality annoyance index. When the overall sound quality annoyance index does not reach the target value, identify the psychoacoustic parameter that has the greatest impact on the overall sound quality annoyance index through correlation analysis. Then, locate the sound source-transmission path combination corresponding to the psychoacoustic parameter based on the energy contribution method of transmission path analysis. Subsequently, with the goal of reducing the adverse effect of the psychoacoustic parameter on the overall sound quality annoyance index, perform secondary optimization on the initial optimization design parameters of the sound source-transmission path combination, and repeat steps S4 to S5. Then return to step S6 until the overall sound quality annoyance index reaches the target value, completing the double closed-loop forward design process.

2. The low-noise forward design method for power-centralized EMU trains combined with sound quality optimization as described in claim 1, characterized in that, The specific implementation process of step S1 is as follows: S1.1 Select no fewer than three power-centralized EMUs of the same type, speed class and good operating condition as sample vehicles, and conduct whole-vehicle line tests on line sections including straight sections, curved sections, bridge sections and tunnel sections. S1.2 Acoustic measurement points were set up inside the passenger compartment of each carriage of the sample vehicle. A total of 6 acoustic measurement points were set up in each carriage, arranged longitudinally along the carriage. The longitudinal positions of the measurement points were set at distances of 0, 1 / 8L, 1 / 4L, 1 / 2L, 3 / 4L, and 7 / 8L from the end of the carriage, respectively, where L is the total longitudinal length of the carriage. Each acoustic measurement point was located on the transverse centerline of the carriage and vertically at a height of 1.6m above the floor to simulate the actual height of the human ear above the ground. The original noise audio signals were simultaneously collected at each measurement point. The signal duration is no less than 30 seconds. After removing transient interference segments at the beginning and end of the original audio signal, a 5-second steady-state segment in the middle is uniformly extracted as the analysis sample. Subsequently, the signal is processed by acoustic software, and invalid data is removed to obtain objective parameter data for each analysis sample. The objective parameter data includes A-weighted sound pressure level and psychoacoustic parameters, including loudness, sharpness, roughness, tone modulation, and fluctuation, to complete the objective test data collection of noise in the passenger compartment of each carriage of the sample vehicle. Then, the A-weighted sound pressure level of each analysis sample is averaged to obtain the measured and verified value of the total A-weighted sound pressure level in the passenger compartment. S1.3 Conduct subjective evaluation tests: Organize no fewer than 30 reviewers aged 18-50, with a male-to-female ratio of 1:1 and normal hearing; use a rating scale to evaluate the subjective annoyance of the analyzed samples; after removing invalid data, obtain the subjective evaluation test results for each analyzed sample, and take the arithmetic mean of the subjective evaluation test results as the benchmark value for subjective annoyance of in-vehicle noise of the existing vehicle models. The specific scoring rules are as follows: The annoyance level is divided into 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, which correspond to the subjective feelings of excellent, very good, very good, good, satisfied, acceptable, generally disliked, very disliked, uneasy, and intolerable, respectively. S1.

4. Based on the vehicle parameters, design speed, formation, car body materials, and sound insulation requirements of the power-centralized EMU, and in conjunction with the measured and verified value of the total A-weighted sound pressure level in the passenger compartment, determine the design target value for the in-car noise sound pressure level. The design target value for the in-car noise sound pressure level is 3-5 dB(A) lower than the measured and verified value of the total A-weighted sound pressure level in the passenger compartment, and is measured in dB(A). At the same time, considering the passengers' requirements for the subjective comfort of the power-centralized EMU, and with reference to the subjective annoyance benchmark value of the in-car noise of the existing vehicle model, determine the target value for the comprehensive annoyance index of sound quality. The target value for the comprehensive annoyance index of sound quality is 1-2 points lower than the subjective annoyance benchmark value of the in-car noise of the existing vehicle model.

3. The low-noise forward design method for power-centralized EMU trains combined with sound quality optimization according to claim 2, characterized in that, The specific implementation process of step S2 is as follows: S2.

1. Establish a vehicle noise simulation model based on statistical energy analysis. The vehicle noise simulation model includes a structural subsystem, an acoustic cavity subsystem, and a sound source subsystem. The structural subsystem consists of the main structural components that make up the body of the power-centralized EMU, the acoustic cavity subsystem consists of the acoustic cavities inside and under the power-centralized EMU, and the sound source subsystem consists of the noise sources that generate noise excitation during the operation of the power-centralized EMU. When dividing the vehicle noise simulation model into subsystems, the power-centralized EMU body structure is divided into 6 structural subsystems: floor, roof, side walls, end walls, window area, and door area; the acoustic cavity is divided into 4 acoustic cavity subsystems: passenger compartment acoustic cavity, driver's cab acoustic cavity, equipment compartment acoustic cavity, and undercarriage space acoustic cavity; and the noise sources are divided into 4 noise source subsystems: wheel-rail noise source, traction equipment noise source, auxiliary equipment noise source, and aerodynamic noise source; among them, the auxiliary equipment noise source is the air conditioning system noise source. The noise transmission relationships between each structural subsystem, each acoustic cavity subsystem, and each sound source subsystem are established. Among them, the vibration generated by the sound source subsystem is transmitted to the passenger compartment through the structural subsystem to characterize the structural sound transmission path; the noise generated by the sound source subsystem is transmitted to the passenger compartment through the acoustic cavity subsystem to characterize the airborne sound transmission path. This characterizes the process of vibration and noise propagating to the passenger compartment through the structural sound transmission path and the airborne sound transmission path, respectively, and thus constructs a vehicle noise simulation model. The input parameters of the whole vehicle noise simulation model were all obtained through experimental testing; among them, the material loss factor was obtained through the damping test method, the sound absorption coefficient was obtained through the standing wave tube method, the structural dynamic characteristic parameters were obtained through modal testing, and the sound source radiated sound power was obtained through the sound intensity method in a semi-anechoic chamber. The above input parameters are imported into the vehicle noise simulation model to obtain the simulation result of the total A-weighted sound pressure level in the passenger compartment. When the deviation between the simulation result of the total A-weighted sound pressure level in the passenger compartment and the measured verification value of the total A-weighted sound pressure level in the passenger compartment obtained in step S1.2 is ≤5dB(A), the vehicle noise simulation model is deemed valid. When the deviation between the simulation result of the total A-weighted sound pressure level in the passenger compartment and the measured verification value of the total A-weighted sound pressure level in the passenger compartment obtained in step S1.2 is >5dB(A), the test is repeated to correct the input parameters of the vehicle noise simulation model until the deviation between the simulation result of the total A-weighted sound pressure level in the passenger compartment and the measured verification value of the total A-weighted sound pressure level in the passenger compartment is ≤5dB(A). Subsequently, based on the valid vehicle noise simulation model, the noise limit allocation and rationality verification of each sound source-transmission path combination are performed. S2.

2. Based on an effective vehicle noise simulation model, obtain the A-weighted sound pressure level contribution value of each sound source-transmission path combination in the total noise value, and determine the initial weight allocation of each sound source-transmission path combination based on the A-weighted sound pressure level contribution value: The sound source-transmission path combination includes at least the following 6 paths, covering more than 95% of the energy contribution sources of in-vehicle noise, specifically: 1) Wheel-rail noise - structural transmission path; 2) Wheel-rail noise - air transmission path; 3) Traction equipment noise - structural transmission path; 4) Traction equipment noise - air transmission path; 5) Aerodynamic noise - air transmission path; 6) Air conditioning system noise - air transmission path. The initial weighting calculation formula for each sound source-transmission path combination is as follows: in, The initial weights are assigned to the j-th sound source-transmission path combination, where n is the total number of sound source-transmission path combinations. The A-weighted sound pressure level contribution value for the j-th sound source-transmission path combination; S2.

3. Based on the initial weight allocation obtained in step S2.2, and considering the coupling correction coefficient... The noise limit for each sound source-transmission path combination is derived by inversely using the principle of energy superposition. ; The formula for superposition of energy is as follows: in, The target value for the sound pressure level of in-vehicle noise; This is the noise limit for the i-th sound source-transmission path combination; It is the coupling correction coefficient for the i-th sound source-transmission path combination, used to correct the nonlinear effect of energy superposition caused by multi-sound source coupling; The larger the value, the greater the actual contribution of the sound source-transmission path combination to the total noise value under the combined action of multiple sound sources. The value ranges from 0.85 to 1.15; Noise limits for each sound source-transmission path combination Based on the initial weight allocation Determine the initial weight allocation. The larger the value, the greater the contribution of the sound source-transmission path combination to the total A-weighted sound pressure level in the guest room; S2.4 Noise limits for each source-transmission path combination obtained by reverse calculation Conduct a rationality verification; set noise limits for each sound source-transmission path combination. The noise limit of each sound source-transmission path combination is compared with the achievable lower limit of the corresponding sound source-transmission path combination to determine whether the noise limit of each sound source-transmission path combination meets the engineering feasibility requirements; when the noise limit of each sound source-transmission path combination... If the noise limits for any sound source-transmission path combination are not lower than the achievable lower noise limit for that combination, then the noise limit allocation scheme is considered reasonable; if the noise limit for any sound source-transmission path combination is lower than the achievable lower noise limit for that combination, then the noise limit allocation scheme is considered reasonable. If the noise level of a sound source-transmission path combination is lower than the lower limit of the achievable noise level for that combination, the noise level of that combination shall be adjusted to the lower limit of the achievable noise level for that combination. The remaining noise level that needs to be reduced after adjustment shall be distributed to other sound source-transmission path combinations until the noise level of each combination meets the feasibility requirements of the project.

4. The low-noise forward design method for power-centralized EMU trains combined with sound quality optimization according to claim 3, characterized in that, The specific implementation process of step S3 is as follows: S3.

1. Select appropriate simulation analysis methods based on the transmission path type and frequency range of each sound source-transmission path combination. Specifically, for sound source-transmission path combinations transmitted through the air, the acoustic boundary element method is used for simulation analysis in the 20Hz~1000Hz frequency band, and the statistical energy analysis method is used for simulation analysis in the 1000Hz~20000Hz frequency band. For sound source-transmission path combinations transmitted through structures, the finite element method is used for simulation analysis in the 20Hz~500Hz frequency band, the transmission path analysis method is used for simulation analysis in the 500Hz~2000Hz frequency band, and the statistical energy analysis method is used for simulation analysis in the 2000Hz~20000Hz frequency band. S3.

2. Based on the simulation analysis method determined in step S3.1, establish independent simulation models for each sound source-transmission path combination. Specifically, for sound source-transmission path combinations transmitted through the air, establish an airborne sound transmission simulation model to simulate the process of sound source noise being transmitted through the air to the passenger compartment. The input parameters of the airborne sound transmission simulation model include at least the sound source power, vehicle body sound insulation, and passenger compartment sound absorption coefficient. For sound source-transmission path combinations transmitted through the structure, establish a structural sound transmission simulation model to simulate the process of sound source vibration excitation being transmitted through the vehicle body structure and forming noise in the passenger compartment. The input parameters of the structural sound transmission simulation model include at least the sound source vibration excitation parameters, vehicle body structural material parameters, and structural damping parameters. The vehicle body structural material parameters include elastic modulus, Poisson's ratio, and density. S3.3 Run the independent simulation model of each sound source-transmission path combination to obtain the A-weighted sound pressure level contribution value of the corresponding sound source-transmission path combination; compare the obtained A-weighted sound pressure level contribution value with the noise limit value corresponding to each sound source-transmission path combination obtained in step S2, and adjust and optimize the key design parameters in each sound source-transmission path combination to obtain the initial optimized design parameters corresponding to each sound source-transmission path combination, subject to the constraint of meeting the corresponding noise limit value.

5. The low-noise forward design method for power-centralized EMU trains combined with sound quality optimization according to claim 4, characterized in that, The specific implementation process of step S4 is as follows: S4.1 Input the initial optimization design parameters corresponding to each sound source-transmission path combination into the vehicle noise simulation model, run the vehicle noise simulation model, and output the total A-weighted sound pressure level of the passenger compartment; The total A-weighted sound pressure level in the passenger compartment is compared with the design target value of the in-vehicle noise sound pressure level. When the deviation between the two is ≤0.5dB(A), the sound pressure level is determined to meet the standard, and the process proceeds to step S5. When the deviation between the two is greater than 0.5 dB(A), the iterative optimization process based on parameter sensitivity ranking is initiated. S4.2 During iterative optimization, parameters should be adjusted in descending order of sensitivity. First, the initial optimization design parameters for each sound source-transmission path combination are changed, and the corresponding changes in vehicle noise are calculated to obtain the sensitivity coefficient of each initial optimization design parameter to vehicle noise. The sensitivity coefficient is used to characterize the degree of influence of changes in the initial optimization design parameters on vehicle noise. The initial optimization design parameters with larger sensitivity coefficients are adjusted first to make the vehicle noise simulation results close to the design target value of the in-vehicle noise sound pressure level. Then, the initial optimization design parameters with smaller sensitivity coefficients are further adjusted to improve the accuracy of parameter adjustment. S4.3 After each round of initial optimization design parameter adjustment, the vehicle noise simulation model is run again to check the total A-weighted sound pressure level in the passenger compartment; the sound pressure level is re-determined to meet the standard; if the requirements are still not met after 5 consecutive iterations, return to step S2 and redistribute the noise limit values ​​for each sound source-transmission path combination.

6. The low-noise forward design method for power-centralized EMU trains combined with sound quality optimization according to claim 5, characterized in that, The multiple linear regression sound quality evaluation and prediction model is expressed as follows: in, The overall annoyance index for sound quality ranges from 1 to 10 points. For loudness, For sharpness, For roughness, For sound scheduling, For volatility; , , , , , The regression coefficients are used to fit and calibrate the subjective evaluation test results using the least squares method.

7. The low-noise forward design method for power-centralized EMU trains combined with sound quality optimization according to claim 6, characterized in that, The specific implementation process of step S6 is as follows: S6.1 Compare the overall annoyance index of sound quality obtained in step S5 with the target value of the overall annoyance index of sound quality; when the overall annoyance index of sound quality reaches the target value of the overall annoyance index of sound quality, the design process ends. When the overall annoyance index of sound quality does not reach the target value, a secondary optimization is initiated. By calculating the correlation coefficient between the psychoacoustic parameters of each analysis sample obtained in step S1.2 and the overall annoyance index M of sound quality, the psychoacoustic parameter with the greatest impact on the overall annoyance index of sound quality is identified. Then, for this psychoacoustic parameter, the energy contribution method based on transmission path analysis is used to calculate the contribution ratio of each sound source-transmission path combination to this psychoacoustic parameter, and the sound source-transmission path combination with the highest contribution ratio is taken as the object of secondary optimization. With the goal of reducing the adverse effects of this psychoacoustic parameter on the overall annoyance index of sound quality, the design parameters of the secondary optimization object are adjusted to obtain the secondary optimized design parameters. S6.2 Update the optimized design parameters to the vehicle noise simulation model, repeat steps S4 to S5 to obtain a new comprehensive annoyance index for sound quality, and then return to step S6.

1. When the comprehensive annoyance index for sound quality reaches the target value of the comprehensive annoyance index for sound quality, the design process ends. If the overall annoyance index of sound quality still does not reach the target value, repeat the second optimization process until the overall annoyance index of sound quality reaches the target value.

Images