Method for quantifying contribution rate of multiple sound sources to factory boundary noise

Abstract

The invention discloses a method for quantifying the contribution rate of multiple sound sources to factory boundary noise, and belongs to the technical field of computer aided design and simulation, and the method comprises the steps: obtaining a sound source feature data set and a monitoring data set; constructing a sound propagation prediction model based on the sound source feature data set and the environment attenuation parameters; simulating an interference effect of sound waves among a plurality of sound sources by adopting a coupling propagation algorithm, and generating coupling sound field data; adjusting parameters of the sound propagation prediction model through a self-adaptive optimization mechanism, enabling the deviation between the coupling sound field data and the monitoring data set to meet a preset precision threshold value, and generating predicted sound field data; and sound source decoupling traceability analysis is performed on the predicted sound field data and the monitoring data set, and the noise contribution rate of each sound source to each measuring point of the factory boundary is calculated. According to the method, the net contribution rate of each sound source to factory boundary noise can be quantified by adopting a computer-aided design means of constructing a coupling sound propagation prediction model, combining an adaptive optimization mechanism for calibration and executing decoupling traceability analysis on a predicted sound field.

Description

A method for quantifying the contribution of multiple sound sources to plant boundary noise. Technical Field

[0001] This invention relates to the field of computer-aided design and simulation technology, and in particular to a method for quantifying the contribution rate of multiple sound sources to plant boundary noise. Background Technology

[0002] In industrial production activities, numerous noise sources within a factory area, such as large machinery, fans, and pumps, generate continuous noise, impacting the factory boundary and surrounding environment. To effectively control noise pollution and meet environmental regulations, it is necessary to identify the contribution of each noise source to noise at specific locations within the factory boundary. This provides a crucial basis for implementing targeted noise reduction measures. Acoustic simulation using computer-aided engineering technology is an important tool for conducting such analyses.

[0003] Among related technologies, Chinese invention patent CN118965764A discloses a noise reduction and intelligent noise management method based on sound source identification and noise analysis, including step S1: constructing a noise distribution simulation calculation model verified by measured noise data in the plant area; step S2: coarse noise source tracing; step S3: noise simulation calculation and noise status assessment under the influence of major noise sources in the entire plant area; step S4: quantitative analysis of the contribution ratio of each sound source to the noise impact at the plant boundary; step S5: determining the noise reduction priority of each sound source according to the order of the contribution ratio of noise at the plant boundary, quantitatively analyzing the noise reduction target value of each sound source, and studying the intelligent noise management scheme for the entire plant area. Step S1 includes: S11: three-dimensional acoustic model construction: the model contains elements that affect noise propagation, including buildings, structures, ground, roads, as well as acoustic boundary conditions with sound absorption and reflection properties, and meteorological environmental information; S12: sound source construction.

[0004] Regarding the aforementioned technologies, the inventors believe they have technical defects in practical applications. First, the sound source construction process heavily relies on preset parameters, failing to establish a dynamic sound source model correlated with the actual operating conditions of the equipment, leading to a disconnect between sound source characteristics and real-world conditions. Second, both the coarse noise source tracing and subsequent simulation analysis ignore the coherence and interference effects of sound waves propagating in space, employing simple energy superposition to calculate sound field superposition, which can cause deviations in industrial scenarios with dense multi-sound-source environments and complex acoustics. Finally, the entire analysis process is an open-loop structure, lacking a closed-loop calibration mechanism for real-time comparison, feedback, and adaptive correction of simulation prediction results with actual plant boundary monitoring data. This results in insufficient accuracy and reliability of the final quantitative assessment of the sound source contribution ratio, making it difficult to support targeted noise reduction decisions. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a method for quantifying the contribution rate of multiple sound sources to factory boundary noise. It employs a computer-aided design approach that involves constructing a coupled sound propagation prediction model, calibrating it using an adaptive optimization mechanism, and then performing decoupling source tracing analysis on the predicted sound field. This method can quantify the net contribution rate of each sound source to factory boundary noise.

[0006] The above objective can be achieved through the following scheme: a method for quantifying the contribution rate of multiple sound sources to factory boundary noise, the method comprising: acquiring acoustic parameters of multiple sound sources within the factory area and monitoring data of multiple measuring points at the factory boundary, generating a sound source feature dataset and a monitoring dataset; constructing a sound propagation prediction model based on the sound source feature dataset and environmental attenuation parameters; simulating the interference effect of sound waves between multiple sound sources using a coupled propagation algorithm based on the sound propagation prediction model, generating coupled sound field data; adjusting the parameters of the sound propagation prediction model through an adaptive optimization mechanism so that the deviation between the coupled sound field data and the monitoring dataset meets a preset accuracy threshold, generating predicted sound field data; performing sound source decoupling and source tracing analysis on the predicted sound field data and the monitoring dataset to trace back and separate the net energy contribution of each sound source, and calculating the noise contribution rate of each sound source to each measuring point at the factory boundary.

[0007] Optionally, the generation of the sound source feature dataset and monitoring dataset includes: measuring the sound power level and spatial location of each sound source using an audio-visual machine to generate basic sound source parameters; collecting equipment vibration data and operating condition change data of each sound source to generate sound source state parameters; combining the basic sound source parameters and the sound source state parameters to establish a storage structure for a dynamic sound source fingerprint to generate a sound source feature dataset; and using high-precision noise monitoring equipment to monitor environmental noise, collecting and recording sound pressure level data at each measuring point within a specific time period to generate a monitoring dataset.

[0008] Optionally, the construction of the sound propagation prediction model includes: determining the atmospheric absorption attenuation coefficient based on local meteorological statistics, and setting initial values ​​for ground effect attenuation and obstacle shielding attenuation based on the factory area's geographical information to generate environmental attenuation parameters; using the dynamic sound source fingerprint as the sound source input, and combining the environmental attenuation parameters to calculate sound wave propagation attenuation, thereby constructing the sound propagation prediction model.

[0009] Optionally, generating coupled sound field data includes: obtaining basic sound field data containing phase information generated by each sound source at the measurement point from the sound propagation prediction model; calculating the interference correction amount based on the basic sound field data and considering the superposition and interaction of sound waves generated by all sound sources at each measurement point; and vector superimposing the basic sound field data and the interference correction amount to generate coupled sound field data.

[0010] Optionally, generating the predicted sound field data includes: comparing the coupled sound field data with the monitoring dataset and calculating the deviation value; when the deviation value exceeds a preset accuracy threshold, automatically identifying the propagation path that causes the deviation to meet the preset conditions; adjusting the attenuation parameter weight of the propagation path and recalculating the sound field distribution until the deviation value meets the preset accuracy threshold, thereby generating the predicted sound field data.

[0011] Optionally, the calculation of the noise contribution rate of each sound source to each measuring point at the plant boundary includes: calculating the equivalent continuous A-weighted sound level of each measuring point based on the monitoring dataset, and generating regional sound energy data in groups; tracing back the sound source energy distribution corresponding to the regional sound energy data according to the predicted sound field data; and removing the energy overlap in the sound source energy distribution through coherence analysis to generate the noise contribution rate of each sound source.

[0012] Optionally, the grouping to generate regional acoustic energy data includes: based on the monitoring dataset, obtaining the sound pressure level measurement value of each measuring point, and calculating the equivalent continuous A-weighted sound level of each measuring point; according to the spatial distribution information of the factory boundary measuring points contained in the monitoring dataset, dividing multiple measuring points into different regional groups; and accumulating the energy of the equivalent continuous A-weighted sound levels corresponding to all measuring points contained in the regional group to generate regional acoustic energy data.

[0013] Optionally, generating the noise contribution rate of each sound source includes: calculating the coherence coefficient between each sound source and identifying the energy overlap portion; subtracting the energy overlap portion from the sound source energy distribution to generate a net sound energy distribution; and calculating the noise contribution rate of each sound source based on the net sound energy distribution.

[0014] Optionally, the method further includes: prioritizing sound sources according to the noise contribution rate to generate a noise reduction strategy; simulating the change in factory boundary noise after implementation using the sound propagation prediction model based on the noise reduction strategy to generate optimization effect data; and dynamically adjusting the noise reduction strategy according to the optimization effect data until the factory boundary noise meets the control standards.

[0015] Based on the same inventive concept, this invention also provides a system for quantifying the noise contribution rate of multiple sound sources to the factory boundary. The system includes: a data acquisition and preprocessing module for acquiring acoustic parameters of multiple sound sources within the factory area and monitoring data from multiple measuring points at the factory boundary, generating a sound source feature dataset and a monitoring dataset; an acoustic model construction module for constructing a sound propagation prediction model based on the sound source feature dataset and environmental attenuation parameters; a sound field coupling simulation module for simulating the interference effect of sound waves between multiple sound sources using a coupling propagation algorithm based on the sound propagation prediction model, generating coupled sound field data; a model adaptive calibration module for adjusting the parameters of the sound propagation prediction model through an adaptive optimization mechanism, so that the deviation between the coupled sound field data and the monitoring dataset meets a preset accuracy threshold, generating predicted sound field data; and a sound source contribution analysis module for performing sound source decoupling and source tracing analysis on the predicted sound field data and the monitoring dataset to trace back and separate the net energy contribution of each sound source, and calculate the noise contribution rate of each sound source to each measuring point at the factory boundary.

[0016] Compared with existing technologies, this invention has the following advantages: By constructing a high-fidelity acoustic digital model and combining it with an adaptive optimization mechanism, this invention enables the model's prediction results to match the on-site monitoring data. This closed-loop simulation method based on real-data calibration improves the accuracy and reliability of sound source contribution rate analysis, ensuring that the analysis results can truly reflect the complex acoustic environment of industrial sites.

[0017] This invention provides a scientific simulation and optimization process for noise reduction strategies. By virtually implementing noise reduction measures and predicting their effects on a validated sound propagation prediction model, different solutions can be evaluated for cost-effectiveness and iteratively optimized before being implemented in actual engineering projects. This allows for the development of a targeted and cost-effective optimal noise reduction strategy, avoiding the blindness and high trial-and-error costs of traditional governance methods.

[0018] This invention establishes a complete technical system from dynamic data acquisition to precise source tracing analysis. By introducing the simulation of dynamic sound source fingerprints and sound wave interference effects, this method can characterize the dynamic characteristics of industrial sound sources as they change with operating conditions and the complex physical phenomena in the sound field. This frees the calculation of noise contribution rate from static and simplified assumptions, providing strong decision support for the refined and scientific management of industrial noise.

[0019] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 is a flowchart illustrating a method for quantifying the contribution rate of multiple sound sources to factory boundary noise according to an embodiment of the present invention.

[0022] Figure 2 is a schematic diagram of the dynamic sound source fingerprint mapping according to an embodiment of the present invention.

[0023] Figure 3 illustrates the adaptive calibration iteration process of the predicted sound field data in an embodiment of the present invention.

[0024] Figure 4 is a bar chart of the noise contribution rate of the sound source in an embodiment of the present invention.

[0025] Figure 5 is a schematic diagram of the structure of a system for quantifying the contribution rate of multiple sound sources to factory boundary noise according to an embodiment of the present invention. Detailed Implementation

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] Referring to Figure 1, an embodiment of the present invention proposes a method for quantifying the contribution rate of multiple sound sources to factory boundary noise. The method employs a computer-aided design approach that involves constructing a coupled sound propagation prediction model and calibrating it using an adaptive optimization mechanism, and then performing decoupling source tracing analysis on the predicted sound field. This approach can quantify the net contribution rate of each sound source to factory boundary noise.

[0028] The method described in this embodiment specifically includes: S1, acquiring acoustic parameters of multiple sound sources within the factory area and monitoring data from multiple measuring points at the factory boundary, generating a sound source feature dataset and a monitoring dataset; optionally, generating the sound source feature dataset and the monitoring dataset includes: measuring the sound power level and spatial location of each sound source using an audio-visual machine to generate basic sound source parameters; collecting equipment vibration data and operating condition change data of each sound source to generate sound source state parameters; combining the basic sound source parameters and the sound source state parameters to establish a storage structure for a dynamic sound source fingerprint, generating a sound source feature dataset; using high-precision noise monitoring equipment to monitor environmental noise, collecting and recording sound pressure level data at each measuring point within a specific time period, generating a monitoring dataset.

[0029] Specifically, the first step is to characterize each potential noise source within the factory area. Operationally, this involves using an acoustic imaging device to measure each sound source. An acoustic imaging device, integrating a microphone array and a camera, uses acoustic beamforming technology to locate the sound source on a visualized image and calculate its acoustic intensity. This measurement provides the spatial location information of each sound source in a three-dimensional Cartesian coordinate system, along with its inherent sound power level; these two parameters together constitute the fundamental parameters of the sound source. However, noise emissions from industrial sources are not constant; they are closely related to their operating conditions. Therefore, it is necessary to simultaneously collect sound source state parameters. This includes installing accelerometers at key equipment components to obtain equipment vibration data and retrieving data reflecting changes in equipment operating load, speed, and throughput from the factory's distributed control system (DCS) or similar monitoring systems. Finally, the static fundamental parameters of the sound source are correlated and fused with the dynamic sound source state parameters to establish a dynamic sound source fingerprint storage structure. This structure is essentially a multidimensional mapping relationship, which maps different combinations of vibration and operating condition data to a specific sound power level and spectral characteristics, forming a time-varying sound source descriptor. Collecting the dynamic sound source fingerprints of all sound sources constitutes a complete sound source feature dataset. As shown in Figure 2, this figure visually illustrates the multidimensional mapping relationship of the sound source dynamic fingerprint. In this figure, the X, Y, and Z axes simulate three key state parameters of the sound source: rotational speed, flow rate, and vibration, respectively. Each scatter point in the figure represents the acoustic fingerprint of the sound source under a specific combination of operating conditions. In parallel, to generate the monitoring dataset, multiple measuring points need to be pre-planned and deployed at the factory boundary. Sound level meters conforming to national Class I or higher standards are installed at these measuring points as high-precision noise monitoring devices. These devices are then activated to perform long-term, continuous, and automated monitoring of environmental noise. The devices collect and record the sound pressure level data of each measuring point within a specific time period according to the set sampling frequency. These raw measurement records, containing timestamps, measuring point locations, and sound pressure level sequences, are then organized and formatted to ultimately form the monitoring dataset.

[0030] For example, consider a high-pressure pump unit in a large chemical plant. First, use an audio-visual device to visualize the pump under typical operating conditions, such as a rated speed of 1500 RPM and a flow rate of... Acoustic measurements were performed on high-pressure pump A, obtaining its spatial coordinates (100m, 20m, 5m) and determining its sound power level to be 105dB, with a peak frequency at 1000Hz. Simultaneously, intelligent accelerometers were installed on the motor and pump body of the high-pressure pump to collect its vibration data in real time. Corresponding high-pressure pump operating parameters, such as flow rate, motor speed, and outlet pressure, were read from the factory's distributed control system (DCS). When the speed of high-pressure pump A increased from 1500RPM to 1800RPM, and the flow rate increased to... At that time, the root mean square value of its vibration acceleration increased from 0.5 grms to 0.7 grms. Synchronous measurement with the acoustic imaging system revealed that the sound power level of high-pressure pump A rose to 108 dB, and the spectral energy at 1500 Hz increased. These correlations, namely "speed 1500 / flow rate 50 / vibration 0.5 grms → sound power level 105 dB / peak 1000 Hz" and "speed 1800 / flow rate 60 / vibration 0.7 grms → sound power level 108 dB / peak 1500 Hz", were stored as the dynamic sound source fingerprint of high-pressure pump A. This process was repeated for all major sound sources within the plant area, ultimately forming a complete sound source characteristic dataset. Simultaneously, to generate the monitoring dataset, three sound level meters conforming to the national Class I standard were installed at measuring point E1 on the east side, measuring point S1 on the south side, and measuring point W1 on the west side of the chemical plant boundary, and were set to sample once per minute. These devices continuously and automatically monitored environmental noise starting at 00:00 every day. For example, at 10:15 AM on a certain day, measuring point E1 recorded a sound pressure level of 58.2 dB(A), measuring point S1 recorded 61.5 dB(A), and measuring point W1 recorded 55.9 dB(A). Half an hour later, at 10:45 AM, these measuring points recorded data again, for example, E1 recorded 59.1 dB(A), S1 recorded 63.0 dB(A), and W1 recorded 56.2 dB(A), and so on continuously. These raw records, with timestamps, geographical coordinates of the measuring points, and continuous sound pressure level values, are systematically processed to form a monitoring dataset for subsequent analysis. This method, by constructing a dynamic sound source fingerprint, surpasses the simplistic treatment of treating the sound source as a constant static source in traditional methods, and characterizes the dynamic acoustic characteristics of industrial sound sources as they change with operating conditions. Simultaneously, the real-time monitoring dataset acquired from high-precision monitoring equipment ensures the ground-based authenticity of model calibration and result verification. This method of collecting the dynamic intrinsic properties of the sound source and the actual noise performance at the factory boundary in a refined and synchronized manner lays the data foundation for the subsequent establishment of a high-fidelity sound propagation prediction model and the realization of accurate quantitative analysis of noise contribution rate, thereby improving the accuracy and reliability of the entire analysis method from the source.

[0031] S2. Based on the sound source feature dataset and environmental attenuation parameters, construct a sound propagation prediction model; optionally, constructing the sound propagation prediction model includes: determining the atmospheric absorption attenuation coefficient according to local meteorological statistics, and setting initial values ​​for ground effect attenuation and obstacle shielding attenuation based on the factory area geographical information, generating environmental attenuation parameters; using the dynamic sound source fingerprint as the sound source input, and combining the environmental attenuation parameters to calculate the sound wave propagation attenuation, constructing a sound propagation prediction model.

[0032] Specifically, firstly, environmental attenuation parameters need to be generated. This process involves quantifying various attenuation mechanisms. Firstly, the atmospheric absorption attenuation coefficient needs to be determined. Air itself is not an ideal propagation medium; its temperature and humidity affect sound energy absorption. By consulting local meteorological statistics to obtain the annual or seasonal average temperature and relative humidity, and using mature calculation methods such as international standard ISO 9613-1, the attenuation of sound waves at different frequencies due to air absorption can be calculated, i.e., the atmospheric absorption attenuation coefficient. Secondly, the initial value for ground effect attenuation needs to be set. When sound waves propagate, in addition to the direct sound reaching the measuring point, there is also reflected sound after being reflected by the ground. The interference between these two will produce enhancement or attenuation effects. Based on the plant's geographical information, such as topographic maps and surface material survey reports, the acoustic characteristics of the ground are preliminarily assessed, and corresponding initial values ​​for ground effect attenuation are set for the sound propagation path. Thirdly, the initial value for obstacle shielding attenuation needs to be set. Buildings, large equipment, and walls within the plant area will all block the direct propagation of sound waves, forming sound shadow zones. Using the 3D model and layout map of buildings contained in the factory's geographical information, the main obstacles between the sound source and the measuring point are identified. Based on acoustic diffraction theory, the shielding attenuation caused by these obstacles is preliminarily estimated. These three types of attenuation parameters together constitute the environmental attenuation parameter set, serving as the basic physical constraints of the model. Next, these environmental attenuation parameters are combined with the established dynamic sound source fingerprint to construct a complete sound propagation prediction model. The core of this model is a mathematical expression used to calculate the sound pressure level generated by any sound source at any measuring point. Specifically, it can be expressed as: ,in, The predicted sound pressure level at the measurement point; It is the sound power level of the sound source, and its value is dynamically provided by the dynamic sound source fingerprint according to the real-time operating conditions. It is geometrical diffusion attenuation, which describes the basic attenuation caused by the divergence of sound energy as distance increases, and is determined by the geometric distance between the sound source and the measuring point; , and These represent the calculated atmospheric absorption attenuation, ground effect attenuation, and obstacle shielding attenuation, respectively. By applying this calculation model to the propagation path between each sound source and each measurement point at the plant boundary, the sound field distribution at the plant boundary under specific operating conditions, when each sound source acts alone, can be predicted.

[0033] For example, regarding the construction of a sound propagation prediction model, taking high-pressure pump A as an example, its sound power level =105dB, the straight-line distance from the specific plant boundary measuring point P is 150m. Calculate the geometric diffusion attenuation based on this distance. The average temperature in the factory area during the measurement period was 20℃, and the relative humidity was 70%. According to ISO 9613-1 standard, the atmospheric absorption attenuation coefficient at 1000Hz is 0.009dB / m. Therefore, the atmospheric absorption attenuation at a distance of 150m... The ground along the sound propagation path is a hard surface. Based on empirical formulas and ground characteristic parameters, the ground effect attenuation is obtained. A 10m high building exists between high-pressure pump A and measuring point P. Using acoustic diffraction theory, calculate the shielding attenuation caused by the building. Substitute these parameters into the model for calculation: The sound propagation prediction model constructed by this method deeply integrates the physical characteristics of the environment and the dynamic behavior of the sound source. Instead of relying on simplified empirical formulas, it uses detailed geographical and meteorological data to perform preliminary physical modeling of key attenuation factors such as atmospheric absorption, ground effect, and obstacle shielding. Simultaneously, by introducing dynamic sound source fingerprints as input, the model can respond to changes in actual factory operating conditions and predict dynamic noise emissions. This modeling approach based on physical principles and dynamic source input provides a highly realistic and structured initial framework for subsequent adaptive model calibration and source tracing, improving the physical fidelity of the model's predictions and its adaptability to complex industrial environments.

[0034] S3. Based on the sound propagation prediction model, a coupled propagation algorithm is used to simulate the interference effect of sound waves between multiple sound sources to generate coupled sound field data; optionally, the generation of coupled sound field data includes: obtaining basic sound field data containing phase information generated by each sound source at the measurement point from the sound propagation prediction model; calculating the interference correction amount based on the basic sound field data and considering the superposition and interaction of sound waves generated by all sound sources at each measurement point; and vector superimposing the basic sound field data and the interference correction amount to generate coupled sound field data.

[0035] Specifically, the first step is to obtain the basic sound field data generated by each sound source at each measurement point at the factory boundary from the already constructed sound propagation prediction model. Unlike calculating sound pressure levels, the basic sound field data here is a complex sound pressure level that includes phase information. For the... The complex sound pressure level of a sound wave generated by a sound source at a certain measuring point Its amplitude can be determined in the following way. The predicted sound pressure level of the sound source at the measuring point can be calculated using a sound propagation prediction model. It is derived from the transformation. Its phase. Then it depends on the length of the sound wave's propagation path. The frequency of sound waves and the initial phase of the sound source A joint decision. Specifically, phase. Primarily determined by the propagation delay term, it can be expressed as ,in It's the speed of sound. Therefore, the basic sound field data... Essentially, it's a vector describing the vibration state of a single-source sound wave at the measurement point. Next, based on the acquired fundamental sound field data from all sound sources, the superposition and interaction of sound waves are calculated using a coupled propagation algorithm. The physical essence of this algorithm is to perform vector superposition of the complex sound pressures from all N sound sources at each measurement point. The total complex sound pressure at the measurement point... The calculation is as follows: ,in, It is the total complex sound pressure of the synthesized sound field at the measuring point; It is the first The fundamental sound field data generated by each sound source at the measuring point is the complex sound pressure level. This vector superposition process automatically includes the interactions between all sound sources, i.e., the interference effect. When two sound waves are close in phase, their amplitudes add up, producing constructive interference; when their phases are opposite, their amplitudes cancel each other out, producing destructive interference. The calculated interference correction is not an independently calculated additional term, but rather the total sound field obtained after the above vector superposition. This demonstrates the difference between the sound field obtained by simply adding energy and the sound field obtained by adding energy. Finally, the result of vector superposition is shown. This is converted into sound pressure levels that can be compared with monitoring data, thus generating coupled sound field data. The total complex sound pressure level is then calculated. By taking the modulus of the signal and applying the standard sound pressure level definition, the coupled sound pressure level that takes into account interference effects can be obtained. Repeat this process for all measurement points at the plant boundary to generate a complete set of distribution data describing the coupled sound field across the entire plant boundary, i.e., coupled sound field data.

[0036] For example, consider high-pressure pump A and another cooling tower B located 180m from the plant boundary measuring point P. The basic acoustic field data of high-pressure pump A at measuring point P is as follows: Its amplitude of 0.005 Pa corresponds to a sound pressure level of 44.65 dB. The basic sound field data of cooling tower B at measuring point P, based on its respective sound propagation prediction model, are as follows: The corresponding sound pressure level is approximately 42 dB. The total complex sound pressure at point P... Pa. Amplitude Pa. Convert this amplitude back to sound pressure level. dB. This method improves the physical fidelity of sound field prediction by introducing phase information and performing vector superposition. It is no longer a coarse estimate based on the simple summation of the energies of each sound source, but rather simulates the interference phenomenon inherent in sound waves as a type of wave. This approach can reproduce noise hotspots or quiet zones generated by interference effects in local areas at the plant boundary, making the prediction results more consistent with the real physical conditions under complex environments. This provides a more robust and reliable sound field physical model for subsequent model calibration and the inversion of sound source contribution rates.

[0037] S4. Adjust the parameters of the sound propagation prediction model through an adaptive optimization mechanism so that the deviation between the coupled sound field data and the monitoring dataset meets a preset accuracy threshold, and generate predicted sound field data; Optionally, generating predicted sound field data includes: comparing the coupled sound field data with the monitoring dataset and calculating the deviation value; when the deviation value exceeds the preset accuracy threshold, automatically identifying the propagation path that causes the deviation that meets the preset conditions; adjusting the attenuation parameter weight of the propagation path and recalculating the sound field distribution until the deviation value meets the preset accuracy threshold, and generating predicted sound field data.

[0038] Specifically, firstly, the coupled sound field data is directly compared with the monitoring dataset to quantify the accuracy of the model prediction. For each boundary measurement point, the difference between the predicted sound pressure level in the coupled sound field data and the corresponding measured sound pressure level in the monitoring dataset is calculated. The differences from all measurement points are then aggregated using a comprehensive index to form the deviation value. The deviation value is calculated using the root mean square error, as shown below: ,in, Represents the overall deviation value; This is the total number of measuring points at the factory boundary; It is in the Predicted sound pressure level in the coupled acoustic field data at each measurement point; It is in the The equivalent continuous A-weighted sound level recorded centrally at each monitoring point. This deviation value. This reflects the average deviation between the current sound propagation prediction model and the real physical world. Next, the calculated deviation value... With a preset accuracy threshold A comparison is then made. This threshold represents the maximum acceptable model error, and its setting is based on the actual engineering accuracy requirements. If Less than or equal to This indicates that the model's prediction accuracy has met the requirements, and the coupled sound field data at this point is adopted as the final predicted sound field data, thus ending the calibration process. If the deviation value... Exceeding the preset accuracy threshold If the predicted value is higher than the measured value, an adaptive optimization mechanism is activated. This mechanism first automatically identifies the propagation paths that cause deviations that meet preset conditions, that is, it identifies the measuring points that contribute the most to the overall deviation and analyzes the key propagation paths affecting these measuring points. For example, for a measuring point whose predicted value is higher than the measured value, it locates the sound sources that contribute the most to its sound level and identifies the propagation paths connecting these sound sources to the measuring point as targets to be optimized. Subsequently, it adjusts the environmental attenuation parameters of these specific propagation paths, especially the ground effect attenuation. and obstacle shielding attenuation The weights are adjusted. This adjustment is not arbitrary, but driven by an optimization algorithm that aims to minimize the deviation. The algorithm determines whether to increase or decrease the attenuation parameter value for a specific path based on the direction and magnitude of the deviation. For example, if the predicted value is too high, the attenuation parameter value on that path is increased. After adjustment, the sound propagation prediction and sound field coupling simulation are re-executed using the updated attenuation parameters to generate a new set of coupled sound field data, and the deviation value is recalculated. This iterative process continues until the deviation value E meets the preset accuracy threshold. The coupled sound field data output by the finally converged model is the high-precision predicted sound field data calibrated with real data. Figure 3 illustrates the iterative process of adaptive calibration of the predicted sound field data.

[0039] For example, considering the adaptive closed-loop calibration process, take the plant boundary measuring point P under the action of high-pressure pump A and cooling tower B as an example. The predicted coupled sound pressure level of measuring point P is 52.5 dB. However, in the actual monitoring dataset, the measured equivalent continuous A-weighted sound level of measuring point P is 48 dB. At this point, the deviation value... The set precision threshold If it is ±2dB, then The model exceeded the allowable error range. The system will automatically identify the overestimation and analyze the main propagation path causing this deviation. The analysis results show that high-pressure pump A contributes a large proportion of the acoustic energy. The system will initiate an adaptive optimization mechanism to fine-tune the environmental attenuation parameters along the path from high-pressure pump A to measuring point P. For example, the system may increase the attenuation caused by obstacles along this path. The pressure was adjusted from 8dB to 12dB. After adjustment, the contribution of high-pressure pump A to measuring point P was... Then, the coupled sound field calculation is performed again. The contribution of cooling tower B remains unchanged, and vector superposition is performed again to obtain a new coupled sound pressure level. Then, the new deviation value is calculated. This process will be repeated iteratively until… The predicted sound field falls within the threshold range of -2dB to 2dB. This method establishes a feedback loop from theoretical models to real-world data, improving the accuracy of sound field prediction. The initially constructed sound propagation prediction model may have systematic biases due to simplified environmental parameters or unknown influencing factors. This method effectively compensates for these biases through repeated comparisons with real monitoring data and adaptive parameter adjustments. It ensures that the final predicted sound field data not only macroscopically matches the overall noise level at the plant boundary but also microscopically captures the details of noise distribution caused by complex propagation effects at specific locations. This provides a sound field snapshot closest to the real physical scenario for subsequent source decoupling and tracing analysis, guaranteeing the reliability and confidence of the contribution rate analysis results.

[0040] S5. Perform source decoupling and source tracing analysis on the predicted sound field data and the monitoring dataset to trace back and separate the net energy contribution of each sound source, and calculate the noise contribution rate of each sound source to each measuring point at the plant boundary.

[0041] Optionally, the calculation of the noise contribution rate of each sound source to each measuring point at the plant boundary includes: calculating the equivalent continuous A-weighted sound level of each measuring point based on the monitoring dataset, and generating regional sound energy data in groups; tracing back the sound source energy distribution corresponding to the regional sound energy data according to the predicted sound field data; and removing the energy overlap in the sound source energy distribution through coherence analysis to generate the noise contribution rate of each sound source.

[0042] Specifically, the first step is to quantify the measured noise levels at the factory boundary. Based on the sound pressure level time series in the monitoring dataset, the equivalent continuous A-weighted sound level at each measuring point within a specified time period is calculated. This is a key indicator representing the average sound energy level during that time period. Subsequently, for ease of macroscopic analysis and management, the factory boundary measuring points can be grouped according to spatial location or management area to generate regional sound energy data. This step is accomplished by summing the equivalent continuous A-weighted sound levels corresponding to all measuring points within each regional group. Since sound pressure level is a logarithmic unit, it must first be converted to a linear quantity such as sound energy or sound intensity during summation. Next, the crucial source tracing step begins. Using highly calibrated predicted sound field data, the aforementioned regional sound energy data or the sound energy of individual measuring points is traced back to explore the underlying sound source energy distribution. Because the predicted sound field data contains every sound source... At each measuring point The resulting sound pressure contribution Measurement points caused by a single point acoustic energy contribution By superimposing the energy contributions of all sound sources to the measuring point, we can obtain the theoretical total energy generated at that measuring point by the combined effect of all sound sources. This energy distribution is the sound source energy distribution. This distribution details the energy flow spectrum from the sound source to the measuring point. However, in a multi-source environment, especially when the sound sources have a certain degree of coherence, the energy of their sound waves during propagation and superposition is not simply added together; there is an energy overlap. Directly using the above sound source energy distribution to calculate the contribution rate will result in a sum exceeding 100% or inaccurate allocation. Therefore, coherence analysis must be used to eliminate these overlaps. Coherence analysis typically involves calculating the correlation coefficients between the sound pressure signals generated by different sound sources at the measuring point. By analyzing these coefficients, the energy gain or attenuation caused by coherent interference can be identified and quantified; this part is the energy overlap. Subtracting this overlap energy from the previously calculated sound source energy distribution yields the net sound energy distribution of each sound source. The net sound energy distribution reflects the independent contribution of each sound source to the sound energy at the measuring point after excluding the influence of interactions with other sound sources. Finally, based on the purified net sound energy distribution, the noise contribution rate of each sound source to each measuring point at the plant boundary is calculated. By performing this calculation on all sound sources and measuring points, a complete and quantified noise contribution rate matrix can be obtained. As shown in Figure 4, this figure uses a stacked bar chart to show the noise contribution rate of multiple sound sources within the plant area to different measuring points at the plant boundary.

[0043] For example, when assessing noise at the boundary of an industrial park, the equivalent continuous A-weighted sound level at each monitoring point within a specified time period is first calculated based on the monitoring dataset. For instance, at monitoring point M1 at the boundary, the measured sound pressure level time series, after processing, yields an equivalent continuous A-weighted sound level of 60.0 dB(A) for one hour. To improve management efficiency, all monitoring points are geographically divided into "North Zone" and "South Zone." The "North Zone" includes monitoring points M1, M2, and M3, with equivalent continuous A-weighted sound levels of 60.0 dB(A), 58.0 dB(A), and 59.0 dB(A), respectively. To generate regional acoustic energy data, these logarithmic sound pressure levels are first converted into linear acoustic energy or sound intensity, and then accumulated. Next, the crucial source tracing step begins. Using highly calibrated predicted sound field data, which shows the sound pressure contribution of the compressor S1 within the factory area at monitoring point M1 alone, its converted acoustic energy contribution is... The acoustic energy contribution of the cooling tower S2 at measuring point M1 alone is: The theoretical total acoustic energy of measuring point M1 is obtained by superimposing the acoustic energy contributions of all sound sources at this measuring point M1. This initially outlines the energy transfer spectrum. However, due to the potential coherence between S1 and S2, their sound waves do not simply add up when superimposed at M1. This method identifies and quantifies the energy overlap caused by coherent interference through coherence analysis, such as calculating the correlation coefficient between the sound pressure signals generated by S1 and S2 at M1. Analysis confirms the existence of an energy overlap at measurement point M1, with a value representing a specific acoustic energy. Subtracting this overlap energy from the previously calculated sound source energy distribution yields the net acoustic energy distribution of each source. Finally, based on the purified net acoustic energy distribution, the noise contribution rate of each source to measurement point M1 is calculated. For example, S1 contributes 25% to M1, and S2 contributes 75%. By performing similar calculations on all sound sources and all measurement points, a complete noise contribution rate matrix can be obtained. This method solves the difficult problem of separating the contribution of individual sound sources from a complex mixed noise field. It transcends the limitations of traditional contribution rate analysis that ignores acoustic coherence. By introducing coherence analysis and an energy overlap elimination mechanism, it ensures the physical rationality of energy conservation and allocation. This decoupled and source-tracing analysis method ensures that the final noise contribution rate is no longer a fuzzy estimate based on a simplified model, but a quantification of the influence of each sound source.

[0044] Optionally, the grouping to generate regional acoustic energy data includes: based on the monitoring dataset, obtaining the sound pressure level measurement value of each measuring point, and calculating the equivalent continuous A-weighted sound level of each measuring point; according to the spatial distribution information of the factory boundary measuring points contained in the monitoring dataset, dividing multiple measuring points into different regional groups; and accumulating the energy of the equivalent continuous A-weighted sound levels corresponding to all measuring points contained in the regional group to generate regional acoustic energy data.

[0045] Specifically, the monitoring dataset first requires basic processing. The dataset contains sound pressure level measurements continuously recorded over time at multiple measurement points at the plant boundary. The first step is to calculate the equivalent continuous A-weighted sound level at each measurement point. This is a widely used indicator in acoustic evaluation; it represents the sound pressure level of a steady-state noise with energy equivalent to fluctuating noise within a specific measurement time T. The calculation formula is: ,in, It is the calculated equivalent continuous A-weighted sound level, in decibels; It is an integral time period, such as one hour or a complete daytime period; Is The instantaneous A-weighted sound pressure level measured at each time point is directly obtained from the monitoring dataset. By performing this calculation on the sound pressure level measurement at each measuring point, the time-varying data stream is transformed into a static value representing the average noise exposure level of that measuring point over a period of time. Next, the core operation of regionalization is carried out, which involves grouping the measuring points according to the spatial distribution information of the measuring points at the plant boundary. This step is not random, but based on actual management needs or physical sound field characteristics. For example, the plant boundary can be divided into four regional groups (east, south, west, and north) based on geographical location; or measuring points near residential areas can be divided into a key focus area group based on the distribution of sensitive points in the surrounding environment. These division criteria are usually stored in the monitoring dataset or are input as configuration information during initialization. Through this operation, the originally discrete multiple measuring points are organized into several regional groups with clear geographical or management significance. Finally, the noise energy within each regional group is summarized to generate regional sound energy data. Since the sound pressure level is logarithmic, it cannot be directly represented by an arithmetic mean or summation. The correct approach is to first calculate the equivalent continuous A-weighted sound level of each measuring point. The acoustic energy is converted back to linear dimensions or a quantity proportional to them, and then accumulated. For a group of regions containing K measurement points, the regional acoustic energy... The calculation is as follows: ,in, The total acoustic energy index represents the total acoustic energy of the region group; it is a dimensionless value that is proportional to the total acoustic energy. It is the first in this region group The equivalent continuous A-weighted sound level at each measuring point. Calculated This refers to the regional acoustic energy data, which reflects the total noise energy load borne by the area.

[0046] For example, a factory has three boundary monitoring points P1, P2, and P3. During a 10-minute monitoring period, the instantaneous sound pressure level at monitoring point P1 stabilizes at 55 dB(A), at monitoring point P2 at 60 dB(A), and at monitoring point P3 at 50 dB(A). Calculate the equivalent continuous A-weighted sound level at each monitoring point. For point P1: Similarly, point P2 is... Point P3 is Based on management requirements, P1 is designated as the "East Region," and P2 and P3 are designated as the "South Region." The third step is to calculate the regional acoustic energy of each region. Eastern region Southern region This process aggregates discrete measurement point data into macroscopic regional acoustic energy data: the acoustic energy index for the eastern region is 316,227.76, and the acoustic energy index for the southern region is 1,100,000, intuitively reflecting that the overall noise load in the southern region is much higher than that in the eastern region, providing managers with a clear basis for decision-making. This method, by integrating point monitoring data into regional acoustic energy data, achieves a macroscopic assessment of the noise situation at the plant boundary. It avoids the one-sidedness caused by focusing on the fluctuations of individual measurement points, instead providing a more stable and representative overall view of the regional noise level. This data aggregation from "point" to "area" not only simplifies the presentation and understanding of the noise situation, enabling managers to quickly identify the plant boundary areas with the most severe noise problems, but also provides target objects for subsequent region-based source analysis and noise reduction strategy formulation. It transforms discrete monitoring information into macroscopic indicators with greater decision-support value, improving the efficiency and targeting of noise management.

[0047] Optionally, generating the noise contribution rate of each sound source includes: calculating the coherence coefficient between each sound source and identifying the energy overlap portion; subtracting the energy overlap portion from the sound source energy distribution to generate a net sound energy distribution; and calculating the noise contribution rate of each sound source based on the net sound energy distribution.

[0048] Specifically, the first step is to identify and quantify the energy overlap between sound sources. At any given measurement point, the total received sound energy is not a simple algebraic sum of the contributions from each source, because the sound waves emitted by different sources may exhibit phase correlation, i.e., coherence. To address this issue, it is necessary to calculate the coherence coefficient between each sound source. For any two sound sources… and The sound pressure signal generated at a certain measuring point and The coherence function between them Defined as: ,in, Represents frequency; yes and The cross-power spectral density describes the degree of correlation between two signals in the frequency domain; and They are and The power spectral density represents the energy distribution of each signal in the frequency domain. It is a time-domain signal simulated using a calibrated sound propagation prediction model, combined with the dynamic sound source fingerprint and propagation path characteristics of the sound source. Coherence function The value ranges between 0 and 1. The closer the value is to 1, the stronger the coherence between the two sound sources at that frequency, and the greater the mutual influence of their energy superposition. By analyzing the entire frequency band, the degree of coherence between any two sound sources can be comprehensively assessed, thereby identifying the energy overlap caused by the coherence effect. Next, the identified energy overlap is subtracted from the preliminary estimated sound source energy distribution to generate the net sound energy distribution. (A sound source...) Total acoustic energy contribution In consideration of all other sound sources After the coherent effect, it can be decomposed into net acoustic energy contributed independently by each component. And the portion of energy overlapping with other sound sources. At a certain measuring point, by The total sound energy generated by each sound source The net acoustic energy distribution can be represented as the sum of the power spectral contributions from each sound source and the cross-power spectral contributions. The generation of the net acoustic energy distribution involves processing the overlapping portion of the energy represented by the cross-power spectrum according to its physical origin. From the total energy contribution of each sound source, the portion generated by its interaction with other sound sources is extracted, retaining its energy contribution as an independent source. This process ultimately yields a net acoustic energy distribution that matches the measured total energy. Finally, based on the calculated net acoustic energy distribution, the noise contribution rate of each sound source can be accurately calculated. For a given measurement point, its total acoustic energy... These are objective measurements determined from monitoring datasets. Sound source. noise contribution rate Defined as the net acoustic energy contribution of the sound source Accounting for the total sound energy at this measuring point Percentage: This calculation ensures that the sum of the contribution rates of all sound sources is exactly 100%, which conforms to the basic physical law of energy conservation. Repeating this calculation for all sound sources and all measurement points or areas of interest yields a detailed noise contribution rate matrix.

[0049] For example, when monitoring noise at the boundary of an industrial park, the equivalent continuous A-weighted sound level at measuring point P1 within a specified time period is first calculated based on the sound pressure level time series, for example, obtaining 65.0 dB(A). For macro-management, P1, along with measuring points P2 and P3 in the same area, can be classified as the "East Zone". Next, using highly calibrated predicted sound field data, the acoustic energy contribution of the two main sound sources—generator A and cooling tower B—to measuring point P1 is analyzed. The data shows that generator A alone contributes 3.0 units of acoustic energy to P1, while cooling tower B alone contributes 7.0 units. These two are then superimposed to obtain the theoretical total energy of measuring point P1. Units. However, due to the coherence between generator A and cooling tower B, coherence analysis is needed to identify and quantify the energy overlap. At measuring point P1, the sound pressure signals generated by generator A and cooling tower B are digitally processed to obtain the cross-power spectral density at a frequency of 1000 Hz. for The self-power spectral density of generator A for The self-power spectral density of cooling tower B for Therefore, its coherence function at that frequency can be calculated. The calculation results show that the two sound sources are completely coherent at a frequency of 1000Hz. Through analysis across the entire frequency band, the energy overlap caused by this coherent effect is identified as 2.0 units of acoustic energy. Subtracting this overlap energy from the theoretical total acoustic energy yields the net acoustic energy distribution. For example, based on the mechanism of coherent overlap, after the energy overlap is proportionally distributed, the net acoustic energy contribution of generator A is 2.0 units, and the net acoustic energy contribution of cooling tower B is 6.0 units. At this time, the measured total acoustic energy at measuring point P1, after energy conversion, is 8.0 units. Finally, based on the net acoustic energy distribution, the noise contribution rate of each sound source is calculated. The contribution rate of generator A to P1 is... The contribution rate of cooling tower B to P1 By repeating this calculation for all sound sources and measurement points, a complete noise contribution rate matrix can be obtained. This method solves the core problem in multi-source contribution rate calculation—the nonlinear superposition of energy—by introducing coherence analysis. It no longer treats sound sources as independent, incoherent entities, but rather reflects their complex interactions as wave sources superimposed in space. By stripping away the overlapping energy components, the resulting net acoustic energy distribution represents the purest and most independent contribution of each sound source, thus ensuring the final calculated noise contribution rate has extremely high physical realism and accuracy.

[0050] Optionally, the method further includes: prioritizing sound sources according to the noise contribution rate to generate a noise reduction strategy; simulating the change in factory boundary noise after implementation using the sound propagation prediction model based on the noise reduction strategy to generate optimization effect data; and dynamically adjusting the noise reduction strategy according to the optimization effect data until the factory boundary noise meets the control standards.

[0051] Specifically, firstly, based on the calculated noise contribution rate of each sound source to each measuring point at the plant boundary, all sound sources are prioritized to generate a preliminary noise reduction strategy. The principle of prioritization is clear: the higher the noise contribution rate of a sound source, the higher its priority for treatment. For example, a list is created, ranking all sound sources from highest to lowest according to their contribution rate to the plant boundary measuring point with the most severe noise exceedance. Sound sources at the top of the list are key sound sources and should be prioritized for treatment. The noise reduction strategy not only includes the priority list of sound sources but can also incorporate cost-benefit analysis, considering the investment required to treat different sound sources versus the expected noise reduction. The noise reduction strategy generated in this way indicates which equipment should be treated first with noise reduction measures such as sound insulation, noise reduction, or vibration reduction. Next, the effects of implementing the noise reduction strategy are virtually simulated using a pre-built and calibrated sound propagation prediction model. This process is called "pre-simulation." Specifically, the sound source parameters selected as the treatment targets are modified at the sound source input end of the model. For example, if the strategy is to install a soundproof enclosure on a compressor, the sound power level can be determined based on the design performance of the enclosure. In the model, reduce the expected insertion loss value, for example, by 20 dB. Then, keeping all other parameters constant, rerun the sound propagation prediction model, including the sound field coupling simulation, to calculate the new predicted sound pressure levels at each measuring point at the plant boundary after implementing the noise reduction measures. Comparing these new predicted sound pressure levels with the sound pressure levels before the measures are implemented yields a set of quantitative optimization effect data, demonstrating the expected improvement in plant boundary noise brought about by the noise reduction strategy. Finally, based on the optimization effect data generated by the simulation, the noise reduction strategy is dynamically adjusted until the expected control target is achieved. If the simulation shows that, after implementing the initial strategy, the noise value at a key measuring point has decreased but still does not meet the national or local plant boundary noise control standards, it is necessary to return to the first step to adjust the strategy. Adjustments may include increasing the treatment of secondary contributing sound sources or using more efficient noise reduction techniques for the main sound sources to achieve a greater noise reduction. For example, if the initial strategy is to reduce noise from source A by 15 dB, but simulation shows that the noise level at the plant boundary still exceeds the standard, the strategy is adjusted to reduce noise from source A by 20 dB, while simultaneously reducing the noise from source B, which has the second highest contribution, by 10 dB. Then, another simulation evaluation is conducted. This cycle of "strategy formulation - simulation preview - evaluation and adjustment" will continue until the simulation results show that the adjusted noise reduction strategy can ensure that the noise levels at all plant boundary measuring points meet the control standards. The final confirmed solution is the optimized, most feasible, and economical final noise reduction strategy.

[0052] For example, the measured noise at measuring point M is 50.0 dB(A), the factory boundary standard is 48.0 dB(A), and the current exceedance is 2.0 dB(A). The contribution rates of each sound source are: A accounts for 52%, B accounts for 38%, and C accounts for 10%. First, a preliminary strategy is generated: since sound source A contributes the most, A is prioritized for treatment. Let the preliminary strategy S1 be to implement measures on sound source A to reduce its sound power level by 5 dB. Next, a virtual simulation is performed. After reducing the sound power level of sound source A by 5 dB in the sound propagation model, the prediction model is rerun, and the simulation result shows that the sound pressure level at measuring point M is 48.8 dB(A). The expected noise reduction is... .but The sound pressure level still exceeded the standard by 0.8 dB(A). Finally, a dynamic adjustment strategy was implemented. Since the standard was not met, further measures were needed. Adjustment strategy S2: More efficient measures were implemented for sound source A to reduce its sound power level by 10 dB, and measures were also taken for sound source B, which contributed the second most, to reduce its sound power level by 3 dB. A simulation was performed again, and after the model ran, the sound pressure level at measurement point M was 46.5 dB(A). The total expected noise reduction was... .at this time This achieves or even surpasses the standards. This closed-loop optimization process ensures that, before actual implementation, a cost-effective and environmentally compliant final noise reduction solution is found through simulation. This method provides a scientific and efficient decision support closed-loop system for complex industrial noise control problems. It transforms abstract contribution rate data into concrete and actionable action plans. This dynamic adjustment and optimization mechanism ensures that the final noise reduction strategy is not only technically effective but also economically reasonable, achieving the goal of achieving the best control effect at the lowest cost and improving the scientific rigor and precision of noise pollution prevention and control in industrial enterprises.

[0053] Based on the same inventive concept, as shown in Figure 5, this invention also provides a system for quantifying the noise contribution rate of multiple sound sources to the factory boundary. The system includes: a data acquisition and preprocessing module for acquiring acoustic parameters of multiple sound sources within the factory area and monitoring data from multiple measuring points at the factory boundary, generating a sound source feature dataset and a monitoring dataset; an acoustic model construction module for constructing a sound propagation prediction model based on the sound source feature dataset and environmental attenuation parameters; a sound field coupling simulation module for simulating the interference effect of sound waves between multiple sound sources using a coupling propagation algorithm based on the sound propagation prediction model, generating coupled sound field data; a model adaptive calibration module for adjusting the parameters of the sound propagation prediction model through an adaptive optimization mechanism, so that the deviation between the coupled sound field data and the monitoring dataset meets a preset accuracy threshold, generating predicted sound field data; and a sound source contribution analysis module for performing sound source decoupling and source tracing analysis on the predicted sound field data and the monitoring dataset to trace back and separate the net energy contribution of each sound source, and calculate the noise contribution rate of each sound source to each measuring point at the factory boundary.

[0054] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any indirect connection method can be applied to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above descriptions are merely exemplary embodiments of the present invention and should not be construed as limiting the scope of the present invention.

[0055] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. A method for quantifying the contribution rate of multiple sound sources to factory boundary noise, characterized in that, The method includes: acquiring acoustic parameters of multiple sound sources within the factory area and monitoring data from multiple measuring points at the factory boundary, generating a sound source feature dataset and a monitoring dataset; constructing a sound propagation prediction model based on the sound source feature dataset and environmental attenuation parameters; simulating the interference effect of sound waves between multiple sound sources using a coupled propagation algorithm based on the sound propagation prediction model, generating coupled sound field data; adjusting the parameters of the sound propagation prediction model through an adaptive optimization mechanism so that the deviation between the coupled sound field data and the monitoring dataset meets a preset accuracy threshold, generating predicted sound field data; performing sound source decoupling and source tracing analysis on the predicted sound field data and the monitoring dataset to trace back and separate the net energy contribution of each sound source, and calculating the noise contribution rate of each sound source to each measuring point at the factory boundary.

2. The method for quantifying the contribution rate of multiple sound sources to factory boundary noise according to claim 1, characterized in that, The generation of the sound source feature dataset and monitoring dataset includes: measuring the sound power level and spatial location of each sound source using an audio-visual machine to generate basic sound source parameters; collecting equipment vibration data and operating condition change data of each sound source to generate sound source state parameters; combining the basic sound source parameters and the sound source state parameters to establish a storage structure for a dynamic sound source fingerprint and generate a sound source feature dataset; and using high-precision noise monitoring equipment to monitor environmental noise, collecting and recording sound pressure level data at each measuring point within a specific time period to generate a monitoring dataset.

3. The method for quantifying the contribution rate of multiple sound sources to factory boundary noise according to claim 2, characterized in that, The construction of the sound propagation prediction model includes: determining the atmospheric absorption attenuation coefficient based on local meteorological statistics, setting initial values ​​for ground effect attenuation and obstacle shielding attenuation based on the factory area's geographical information, and generating environmental attenuation parameters; using the dynamic sound source fingerprint as the sound source input, and combining the environmental attenuation parameters to calculate the sound wave propagation attenuation, thereby constructing the sound propagation prediction model.

4. The method for quantifying the contribution rate of multiple sound sources to factory boundary noise according to claim 3, characterized in that, The process of generating coupled sound field data includes: obtaining basic sound field data containing phase information generated by each sound source at the measurement point from the sound propagation prediction model; calculating the interference correction amount based on the basic sound field data and considering the superposition and interaction of sound waves generated by all sound sources at each measurement point; and vector superimposing the basic sound field data and the interference correction amount to generate coupled sound field data.

5. The method for quantifying the contribution rate of multiple sound sources to factory boundary noise according to claim 4, characterized in that, The process of generating predicted sound field data includes: comparing the coupled sound field data with the monitoring dataset and calculating the deviation value; when the deviation value exceeds a preset accuracy threshold, automatically identifying the propagation path that causes the deviation to meet the preset conditions; adjusting the attenuation parameter weight of the propagation path and recalculating the sound field distribution until the deviation value meets the preset accuracy threshold, thereby generating predicted sound field data.

6. The method for quantifying the contribution rate of multiple sound sources to factory boundary noise according to claim 1, characterized in that, The calculation of the noise contribution rate of each sound source to each measuring point at the plant boundary includes: calculating the equivalent continuous A-weighted sound level of each measuring point based on the monitoring dataset, and generating regional sound energy data in groups; tracing back the sound source energy distribution corresponding to the regional sound energy data according to the predicted sound field data; and removing the energy overlap in the sound source energy distribution through coherence analysis to generate the noise contribution rate of each sound source.

7. The method for quantifying the contribution rate of multiple sound sources to factory boundary noise according to claim 6, characterized in that, The process of generating regional acoustic energy data by grouping includes: acquiring sound pressure level measurements at each measuring point based on the monitoring dataset, and calculating the equivalent continuous A-weighted sound level at each measuring point; dividing multiple measuring points into different regional groups according to the spatial distribution information of the factory boundary measuring points contained in the monitoring dataset; and accumulating the energy of the equivalent continuous A-weighted sound levels corresponding to all measuring points within the regional group to generate regional acoustic energy data.

8. The method for quantifying the contribution rate of multiple sound sources to factory boundary noise according to claim 6, characterized in that, The process of generating the noise contribution rate of each sound source includes: calculating the coherence coefficient between each sound source and identifying the energy overlap portion; subtracting the energy overlap portion from the sound source energy distribution to generate a net sound energy distribution; and calculating the noise contribution rate of each sound source based on the net sound energy distribution.

9. The method for quantifying the contribution rate of multiple sound sources to factory boundary noise according to claim 1, characterized in that, The method further includes: prioritizing sound sources according to the noise contribution rate to generate a noise reduction strategy; simulating the change in factory boundary noise after implementation using the sound propagation prediction model based on the noise reduction strategy to generate optimization effect data; and dynamically adjusting the noise reduction strategy according to the optimization effect data until the factory boundary noise meets the control standards.

10. A system for quantifying the contribution rate of multiple sound sources to factory boundary noise, characterized in that, The system includes: a data acquisition and preprocessing module for acquiring acoustic parameters of multiple sound sources within the factory area and monitoring data from multiple measuring points at the factory boundary, generating a sound source feature dataset and a monitoring dataset; an acoustic model construction module for constructing a sound propagation prediction model based on the sound source feature dataset and environmental attenuation parameters; a sound field coupling simulation module for simulating the interference effect of sound waves between multiple sound sources using a coupling propagation algorithm based on the sound propagation prediction model, generating coupled sound field data; a model adaptive calibration module for adjusting the parameters of the sound propagation prediction model through an adaptive optimization mechanism, ensuring that the deviation between the coupled sound field data and the monitoring dataset meets a preset accuracy threshold, generating predicted sound field data; and a sound source contribution analysis module for performing sound source decoupling and source tracing analysis on the predicted sound field data and the monitoring dataset, to trace back and separate the net energy contribution of each sound source, and calculate the noise contribution rate of each sound source to each measuring point at the factory boundary.

Images