Omnibearing sound wave denoising system and self-adaptive depth adjusting method thereof

Abstract

The invention discloses an omni-directional sound wave de-noising system and a self-adaptive depth adjusting method thereof. The method comprises the following steps: installing sound wave emitters at four corners and a plurality of preset positions of a target space; the noise level, acoustic characteristics and sound source distribution in the target space are monitored in real time through an acoustic sensor, and acoustic environment data are generated; calculating a spatial position, a propagation path and energy distribution of a noise source by using a time difference positioning algorithm and a beam forming algorithm, and constructing a three-dimensional sound field model; analyzing the three-dimensional sound field model through a ResNet50 network model, identifying a noise type, and generating a de-noising strategy corresponding to the noise type; the vibration mode, the phase and the amplitude of each sound wave emitter are regulated and controlled by using a self-adaptive wave beam adjustment algorithm, and noise elimination waves with specific directions and shapes are generated. Through the three-dimensional sound field modeling and self-adaptive beam adjustment method, all-directional noise elimination is realized, and the method has the advantages of real-time dynamic adjustment, accurate classification and high noise reduction effect.

Description

Technical Field

[0001] The present invention relates to the technical field of sound wave noise reduction, and in particular to an omnidirectional sound wave noise reduction system and an adaptive depth adjustment method thereof. Background Art

[0002] With the development of social economy and the acceleration of urbanization, noise pollution has become one of the important issues affecting people's lives, work and health. There are many types of noise sources, including noise from industrial equipment operation, noise from vehicles, environmental noise in public places and various noises in family life. These noises not only interfere with people's daily communication and rest, but also cause hearing loss, mental stress, sleep disorders and other health problems. Therefore, how to effectively reduce the impact of noise and improve the indoor and outdoor sound environment has become one of the important research directions of current noise control technology.

[0003] Existing noise control technologies mainly include passive noise reduction and active noise reduction. Passive noise reduction mainly relies on sound insulation materials, sound-absorbing structures and damping designs to absorb or isolate noise through physical means. However, passive noise reduction technology has limited effect on low-frequency noise control, and is limited by material structure and spatial conditions, and cannot meet the full-band noise reduction requirements in complex noise environments. In addition, passive noise reduction systems are usually static and cannot be adjusted in real time according to the dynamic changes of noise sources, and have poor flexibility.

[0004] As an effective supplement to passive noise reduction technology, active noise reduction technology achieves noise cancellation by emitting sound waves with opposite phases to the noise signal, especially showing significant advantages in low-frequency noise control. At present, active noise reduction technology is widely used in specific scenarios such as headphones, cars, and aircraft cockpits, but its application scope is mainly limited to local spaces. Most existing active noise reduction systems rely on single-point or local sensors for noise collection and feedback control, resulting in limited noise reduction coverage and inability to achieve all-round noise control in larger spaces. At the same time, the sound wave transmitters of existing systems usually use fixed parameter settings, lack adaptive adjustment capabilities, and are difficult to cope with complex acoustic environments and dynamic changes in noise sources.

[0005] In the prior art, some systems attempt to model and locate the noise field through microphone arrays. However, these methods are usually limited to simple two-dimensional plane modeling and cannot accurately reflect the noise propagation path and energy distribution in three-dimensional space. In addition, the spatial positioning and tracking of noise sources rely on traditional time difference positioning algorithms and simple filter control, with low positioning accuracy and noise cancellation effects. Especially when the noise source is reflected, refracted or propagated in multiple paths, the system is prone to positioning errors, resulting in unsatisfactory noise cancellation effects.

[0006] For the identification of noise types, most existing technologies rely on feature extraction and traditional classification algorithms, such as classifiers based on statistical features. However, these methods are difficult to achieve high-precision classification in complex noise environments, especially for the distinction between steady-state noise, transient noise and periodic noise, and cannot provide accurate input data for subsequent noise cancellation strategies. In addition, when implementing noise cancellation strategies, existing active noise reduction systems mostly use beamforming methods with fixed parameters, which cannot adaptively adjust the emission direction, amplitude and phase of sound waves according to the actual environment, resulting in difficulty in achieving accurate phase matching between the noise cancellation wave and the target noise.

[0007] In actual application scenarios, the acoustic environment of indoor spaces often has dynamic characteristics. Environmental factors such as temperature, humidity, air pressure and vibration will affect the propagation speed and path of sound waves. Most existing active noise reduction technologies do not take these environmental dynamic factors into consideration and lack the ability to adjust sound wave parameters in real time, resulting in a decrease in the noise reduction performance of the system in complex environments. In addition, existing technologies are relatively simple in terms of system feedback control and parameter optimization, usually relying only on fixed feedback mechanisms, and failing to combine real-time acoustic data for dynamic parameter optimization, making it difficult to achieve efficient and stable noise reduction effects.

[0008] In summary, the existing active noise reduction technology has defects such as limited coverage, low noise type recognition accuracy, insufficient adaptive adjustment capability and poor environmental dynamic adaptability, which makes it difficult to meet the needs of efficient noise control in all aspects and multiple scenarios. Summary of the invention

[0009] One object of the present invention is to propose an all-round sound wave noise cancellation adaptive depth adjustment method. The present invention overcomes the shortcomings of the prior art and realizes all-round, high-precision, and dynamically adaptive noise control by deploying multiple sound wave transmitters in the target space, combining real-time acoustic environment data acquisition, three-dimensional sound field modeling, deep learning noise identification, and adaptive beam adjustment technology. The present invention can monitor the spatial position, propagation path, and energy distribution of the noise source in real time, accurately identify the noise type using a deep learning model, and generate a corresponding noise cancellation strategy based on the classification results. The emission parameters of the sound wave are dynamically controlled through an adaptive beam adjustment algorithm to ensure that the phase of the noise cancellation wave and the target noise matches, thereby achieving an efficient noise cancellation effect.

[0010] A method for adaptive depth adjustment of omnidirectional sound wave noise cancellation according to an embodiment of the present invention comprises the following steps:

[0011] S1. Install acoustic wave transmitters at the four corners of the target space and multiple preset positions;

[0012] S2, monitoring the noise level, acoustic characteristics and sound source distribution in the target space in real time through the acoustic sensor built into the sound wave transmitter, and generating acoustic environment data containing noise time domain, frequency domain and spatial information;

[0013] S3. Based on the acoustic environment data, the spatial position, propagation path and energy distribution of the noise source are calculated using the time difference positioning algorithm and the beamforming algorithm to construct a three-dimensional sound field model;

[0014] S4. Analyze the three-dimensional sound field model through the ResNet50 network model, identify the noise type, and generate a noise reduction strategy corresponding to the noise type;

[0015] S5. Based on the noise cancellation strategy, an adaptive beam adjustment algorithm is used to control the vibration mode, phase and amplitude of each sound wave transmitter to generate a noise cancellation wave with a specific direction and shape, and to achieve phase cancellation between the noise cancellation wave and the target noise.

[0016] Optionally, the S3 specifically includes:

[0017] S31. Use a time difference positioning algorithm to calculate the spatial position of the noise source, using the time difference of noise signals received by different microphones in the microphone array of the acoustic sensor:

[0018]

[0019] Among them, Δt ij It represents the time difference between microphone i and microphone j receiving the noise signal, t i and t j They represent the time when the corresponding microphone receives the signal, d i and d j Respectively represent the distances from the noise source to microphone i and microphone j, and c represents the speed of sound;

[0020] S32, combining the time difference positioning result of the noise source with the spatial coordinate data, and using the geometric triangulation positioning method to calculate the specific position of the noise source in the three-dimensional space:

[0021]

[0022] Among them, (x s ,y s ,z s ) represents the three-dimensional spatial coordinates of the noise source, (x i ,y i ,z i ) represents the coordinate position of microphone i, t i represents the time it takes for the noise signal to propagate to microphone i, and N represents the number of microphones;

[0023] S33. Based on the three-dimensional positioning result of the noise source, calculate the energy distribution of the noise source in the target space:

[0024]

[0025] Where P(f) represents power spectrum density, X(f,t) represents frequency domain data, and T represents time period;

[0026] S34, modeling the propagation path of the noise signal through a beamforming algorithm to determine the direction of noise propagation and the main reflection path;

[0027] S35. Construct a three-dimensional sound field model in the target space according to the spatial coordinates, propagation path and energy distribution of the noise source, wherein the three-dimensional sound field model includes the positioning coordinates of the noise source, the noise intensity distribution and the noise propagation path data.

[0028] Optionally, the S4 specifically includes:

[0029] S41, inputting data into the three-dimensional sound field model to obtain the spatial position, propagation path, energy distribution and noise intensity data of the noise source;

[0030] S42, extracting time domain features and spectrum features of the noise signal, wherein the time domain features include peak value, mean value and pulse amplitude;

[0031] S43, inputting the extracted noise signal features into the ResNet50 network model, and classifying the noise types using the time domain features and spectrum features of the noise, wherein the noise types include steady-state noise, transient noise, and periodic noise;

[0032] S44. Generate a noise reduction strategy corresponding to the noise type according to the classification result, wherein the noise reduction strategy includes a noise reduction target frequency band, a beam direction, and a vibrator parameter setting.

[0033] Optionally, the S5 specifically includes:

[0034] S51, according to the noise cancellation strategy, obtain the type of noise, target frequency band, sound field distribution and noise propagation path, and determine the emission parameters of each sound wave transmitter, wherein the emission parameters include the vibration mode, amplitude, phase and emission direction of the vibrator;

[0035] S52, controlling the vibrator arrays of multiple acoustic wave transmitters by an adaptive beam adjustment algorithm:

[0036]

[0037] Among them, y(t) represents the output signal of the acoustic wave transmitter, w i represents the control weight of the vibrator, x i(t) represents the input signal of the i-th vibrator, and N represents the number of vibrators;

[0038] S53, adjusting the vibration amplitude and phase of each vibrator, and controlling the directivity of the sound wave by a beamforming method;

[0039]

[0040] Among them, w opt represents the optimal weight, d(t) represents the expected denoised signal, and E represents the mathematical expectation;

[0041] S54, optimizing the emission direction of the beam based on the characteristics of noise propagation in the target space;

[0042]

[0043] Among them, P(θ) represents the acoustic wave transmission power, θ represents the transmission direction angle of the beam, and d i represents the distance from the ith vibrator to the reference point, and k represents the wave number;

[0044] S55. Combined with the real-time propagation path and energy distribution data of noise, dynamically adjust the phase and amplitude of each sound wave transmitter to generate noise-cancelling waves in a specific direction and shape.

[0045] Optionally, the sound wave transmitter is composed of a plurality of small vibrators.

[0046] An omnidirectional sound wave noise reduction system according to an embodiment of the present invention includes:

[0047] Multiple acoustic wave transmitters for transmitting noise-cancelling waves and collecting acoustic environment data;

[0048] Data acquisition module, used to collect time domain, frequency domain and spatial information of noise;

[0049] A data processing module is used to analyze and process the monitored data;

[0050] A noise classification module, used to classify noise types;

[0051] The adaptive beam adjustment module is used to adjust the vibration mode, phase and transmission power of the acoustic wave transmitter according to the noise cancellation strategy.

[0052] The beneficial effects of the present invention are:

[0053] First, the present invention achieves full coverage of the entire space by installing sound wave transmitters at the four corners of the target space and multiple preset positions, breaking through the limitations of local noise reduction in the prior art, expanding the noise reduction range, and achieving precise noise control in a larger space. Through the built-in acoustic sensor of the sound wave transmitter, the present invention can collect environmental data such as noise level, acoustic characteristics and sound source distribution in the target space in real time, ensuring comprehensive monitoring of noise and providing basic data support for subsequent sound field modeling and strategy optimization.

[0054] Secondly, the present invention uses the time difference positioning algorithm and the beamforming algorithm, combined with three-dimensional spatial data, to accurately calculate the spatial position, propagation path and energy distribution of the noise source, and construct a three-dimensional sound field model, thereby solving the defects of the prior art that the two-dimensional sound field modeling has low accuracy and cannot accurately reflect the noise propagation path. By introducing the ResNet50 deep learning network model, the present invention can efficiently analyze the time domain characteristics and spectral characteristics of the noise signal, realize the accurate classification of steady-state noise, transient noise and periodic noise, and generate targeted noise reduction strategies based on the classification results. This technological breakthrough significantly improves the accuracy of noise classification, enabling the system to flexibly adjust the noise reduction parameters according to different types of noise sources, thereby optimizing the noise reduction effect.

[0055] In addition, based on the noise cancellation strategy, the present invention dynamically controls the vibration mode, amplitude, phase and emission direction of each sound wave transmitter through an adaptive beam adjustment algorithm to form a noise cancellation wave of a specific direction and shape. Unlike the existing fixed parameter beamforming method, the present invention can achieve real-time dynamic adjustment of the sound wave to ensure that the noise cancellation wave and the target noise are accurately offset in phase, thereby effectively eliminating the noise. The present invention also combines real-time environmental dynamic data, including air pressure, temperature, humidity and vibration factors, to dynamically optimize the delay, phase matching and transmission power of the sound wave, further improving the adaptability and stability of the system in complex dynamic environments.

[0056] In summary, the present invention realizes all-round, high-precision, and adaptive noise control through all-round, multi-dimensional acoustic data collection, three-dimensional sound field modeling, deep learning noise classification, adaptive beam adjustment and other technical means, and can provide stable and efficient noise reduction effects in dynamic and complex acoustic environments. Compared with the prior art, the present invention not only expands the noise reduction coverage, improves the noise positioning accuracy and noise classification accuracy, but also has dynamic adaptive adjustment capabilities, can perform real-time parameter optimization according to environmental changes, and is suitable for a variety of application scenarios such as industry, transportation, and home, creating a quieter and more comfortable sound environment for users. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0058] Figure 1 A flow chart of an all-round acoustic noise cancellation adaptive depth adjustment method proposed by the present invention;

[0059] Figure 2 This is a schematic diagram of an omnidirectional sound wave noise reduction system proposed by the present invention. DETAILED DESCRIPTION

[0060] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0061] refer to Figure 1 and Figure 2 , a method for adaptive depth adjustment of omnidirectional acoustic noise cancellation, comprising the following steps:

[0062] S1. Install acoustic wave transmitters at the four corners of the target space and multiple preset positions;

[0063] S2, monitoring the noise level, acoustic characteristics and sound source distribution in the target space in real time through the acoustic sensor built into the sound wave transmitter, and generating acoustic environment data containing noise time domain, frequency domain and spatial information;

[0064] S3. Based on the acoustic environment data, the spatial position, propagation path and energy distribution of the noise source are calculated using the time difference positioning algorithm and the beamforming algorithm to construct a three-dimensional sound field model;

[0065] S4. Analyze the three-dimensional sound field model through the ResNet50 network model, identify the noise type, and generate a noise reduction strategy corresponding to the noise type;

[0066] S5. Based on the noise cancellation strategy, an adaptive beam adjustment algorithm is used to control the vibration mode, phase and amplitude of each sound wave transmitter to generate a noise cancellation wave with a specific direction and shape, and to achieve phase cancellation between the noise cancellation wave and the target noise.

[0067] In this implementation, S3 specifically includes:

[0068] S31. Use a time difference positioning algorithm to calculate the spatial position of the noise source, using the time difference of noise signals received by different microphones in the microphone array of the acoustic sensor:

[0069]

[0070] Among them, Δt ij It represents the time difference between microphone i and microphone j receiving the noise signal, t i and t j They represent the time when the corresponding microphone receives the signal, d i and d j Respectively represent the distances from the noise source to microphone i and microphone j, and c represents the speed of sound;

[0071] S32, combining the time difference positioning result of the noise source with the spatial coordinate data, and using the geometric triangulation positioning method to calculate the specific position of the noise source in the three-dimensional space:

[0072]

[0073] Among them, (x s ,y s ,z s ) represents the three-dimensional spatial coordinates of the noise source, (x i ,y i ,z i ) represents the coordinate position of microphone i, t i represents the time it takes for the noise signal to propagate to microphone i, and N represents the number of microphones;

[0074] S33. Based on the three-dimensional positioning result of the noise source, calculate the energy distribution of the noise source in the target space:

[0075]

[0076] Where P(f) represents power spectrum density, X(f,t) represents frequency domain data, and T represents time period;

[0077] S34, modeling the propagation path of the noise signal through a beamforming algorithm to determine the direction of noise propagation and the main reflection path;

[0078] S35. Construct a three-dimensional sound field model in the target space according to the spatial coordinates, propagation path and energy distribution of the noise source, wherein the three-dimensional sound field model includes the positioning coordinates of the noise source, the noise intensity distribution and the noise propagation path data.

[0079] In this implementation manner, the S4 specifically includes:

[0080] S41, inputting data into the three-dimensional sound field model to obtain the spatial position, propagation path, energy distribution and noise intensity data of the noise source;

[0081] S42, extracting time domain features and spectrum features of the noise signal, wherein the time domain features include peak value, mean value and pulse amplitude;

[0082] S43, inputting the extracted noise signal features into the ResNet50 network model, and classifying the noise types using the time domain features and spectrum features of the noise, wherein the noise types include steady-state noise, transient noise, and periodic noise;

[0083] S44. Generate a noise reduction strategy corresponding to the noise type according to the classification result, wherein the noise reduction strategy includes a noise reduction target frequency band, a beam direction, and a vibrator parameter setting.

[0084] In this implementation manner, S5 specifically includes:

[0085] S51, according to the noise cancellation strategy, obtain the type of noise, target frequency band, sound field distribution and noise propagation path, and determine the emission parameters of each sound wave transmitter, wherein the emission parameters include the vibration mode, amplitude, phase and emission direction of the vibrator;

[0086] S52, controlling the vibrator arrays of multiple acoustic wave transmitters by an adaptive beam adjustment algorithm:

[0087]

[0088] Among them, y(t) represents the output signal of the acoustic wave transmitter, w i represents the control weight of the vibrator, x i (t) represents the input signal of the i-th vibrator, and N represents the number of vibrators;

[0089] S53, adjusting the vibration amplitude and phase of each vibrator, and controlling the directivity of the sound wave by a beamforming method;

[0090]

[0091] Among them, w opt represents the optimal weight, d(t) represents the expected denoised signal, and E represents the mathematical expectation;

[0092] S54, optimizing the emission direction of the beam based on the characteristics of noise propagation in the target space;

[0093]

[0094] Among them, P(θ) represents the acoustic wave transmission power, θ represents the transmission direction angle of the beam, and d i represents the distance from the ith vibrator to the reference point, and k represents the wave number;

[0095] S55. Combined with the real-time propagation path and energy distribution data of noise, dynamically adjust the phase and amplitude of each sound wave transmitter to generate noise-cancelling waves in a specific direction and shape.

[0096] In this embodiment, the sound wave transmitter is composed of a plurality of small vibrators.

[0097] An all-round sound wave noise cancellation system, comprising:

[0098] Multiple acoustic wave transmitters for transmitting noise-cancelling waves and collecting acoustic environment data;

[0099] Data acquisition module, used to collect time domain, frequency domain and spatial information of noise;

[0100] A data processing module is used to analyze and process the monitored data;

[0101] A noise classification module, used to classify noise types;

[0102] The adaptive beam adjustment module is used to adjust the vibration mode, phase and transmission power of the acoustic wave transmitter according to the noise cancellation strategy.

[0103] Embodiment 1:

[0104] In order to verify the actual application effect of "an all-round sound wave noise reduction system and its adaptive depth adjustment method" described in the present invention, this embodiment selects an open office space of 300 square meters as a test scene. There are about 60 office workers in this space, and the noise sources mainly include the operating sound of office equipment (such as printers, computer fan sounds), the conversations of the crowd, and external traffic noise. Especially during peak office hours, the dense conversations of the crowd and the equipment noise are superimposed on each other, and the noise value can reach up to 78dB, which is much higher than the national office noise standard (≤55dB). The goal of the present invention is to achieve all-round noise reduction in this complex acoustic environment by implementing the method, reduce the noise level in the office area, and solve the problems of limited coverage of existing passive noise reduction technology, poor low-frequency noise processing effect, and insufficient noise adaptive adjustment ability.

[0105] In this office space, first, according to the method of the present invention, 16 sound wave transmitters are installed in the four corners of the space and multiple preset positions. The positions of these transmitters are precisely calculated to ensure that the sound waves can achieve all-round coverage of the entire office space. The acoustic sensors built into the sound wave transmitters monitor the acoustic environment in real time, collect the time domain, frequency domain and spatial distribution information of the noise, and transmit it to the data processing system. The system uses the time difference positioning algorithm and the beamforming algorithm to construct a three-dimensional sound field model, and calculate the location, propagation path and energy distribution of the noise source.

[0106] The system analyzes the sound field model data through the ResNet50 deep learning network model, and divides the noise into steady-state noise (printer, equipment operation sound, average 58dB), transient noise (object falling sound, peak 80dB) and periodic noise (air conditioner operation sound, continuous 61dB). Based on the classification results, the system generates targeted noise reduction strategies and sets corresponding frequency bands and emission parameters for different types of noise.

[0107] The system uses an adaptive beam adjustment algorithm to adjust the vibration mode, phase, amplitude and emission direction of the acoustic wave transmitter in real time to generate noise-canceling waves with specific directions and shapes. These noise-canceling waves are in the opposite phase to the target noise, thus achieving phase cancellation of the noise.

[0108] Table 1. Experimental data comparison of all-round acoustic noise cancellation system

[0109]

[0110] It can be seen from the above experimental data comparison table that the omnidirectional sound wave noise reduction system described in the present invention shows significant noise reduction effects in different types of noise environments, especially in complex and dynamic acoustic environments, showing its real-time, high precision and stability.

[0111] For steady-state noise (such as continuous background noise generated by the operation of equipment such as air conditioners, printers, and computer fans), the noise level before noise reduction was between 58dB and 61dB, and the level after noise reduction was significantly reduced to 46dB to 48dB, with an average noise reduction of 12 to 13dB. This result shows that the system can effectively capture the propagation path and energy distribution of steady-state noise through three-dimensional sound field modeling and adaptive beam adjustment algorithm, and generate targeted inverse phase sound waves to accurately offset these continuous noises. In addition, by identifying and classifying the features of steady-state noise through a deep learning network model, the system avoids the problems of "over-noise reduction" or "parameter mismatch" in traditional noise reduction methods, ensuring the accuracy and continuity of the noise reduction process, and greatly improving the comfort of the office environment.

[0112] For transient noise (sudden noises such as falling objects, dragging chairs, and clapping documents), this type of noise is characterized by short duration, large amplitude, and strong impact, which often causes instantaneous interference to users, seriously affecting work efficiency and experience. During the test, the peak value of transient noise before noise reduction reached 75dB to 80dB, and after being processed by the system of the present invention, the noise level was reduced to 59dB to 62dB, and the average noise reduction amplitude reached 16 to 18dB. The efficient feedback mechanism and adaptive beam adjustment algorithm of this system play a key role. The system can respond and complete sound wave control in about 12ms, ensuring that the noise reduction wave and the transient noise are phase matched in a very short time, and the error is controlled within 2°. This performance advantage is difficult to achieve in traditional noise reduction technology, which further proves the real-time and accuracy of this system in processing high-energy burst noise.

[0113] For periodic noise (such as the operating noise of air conditioners or fans), it is characterized by changes in noise intensity within a certain period. The noise level before noise reduction is about 63dB, and after system processing, it drops to 50dB, and the noise reduction reaches 13dB. The elimination of periodic noise mainly depends on the three-dimensional sound field modeling technology and real-time beam control in the present invention. The system can accurately capture the spectral characteristics and propagation path of periodic noise, and dynamically adjust the emission direction, amplitude and phase of the sound wave to cope with the characteristics of noise changing over time. At the same time, the system combines environmental dynamic data (such as temperature, humidity and air pressure) to adaptively adjust the noise reduction parameters, further ensuring the stability and continuity of periodic noise elimination.

[0114] For external interference noise (such as road traffic noise, vehicle horns, construction noise and other complex external noise), the noise level before system processing is as high as 78dB, and after noise reduction, it is effectively reduced to 58dB, with a noise reduction of 20dB. External noise has strong uncertainty, complex propagation paths and is greatly affected by changes in the external environment. This system can quickly identify the location of the noise source, propagation path and noise type through efficient three-dimensional sound field modeling algorithms and ResNet50 deep learning networks, and dynamically generate accurate noise reduction strategies based on different noise characteristics. In addition, the adaptive beam adjustment algorithm ensures that the output parameters of the sound wave transmitter match the external noise in real time, so that the system can still maintain a high noise reduction effect and stability in an external noise environment.

[0115] In addition, from the perspective of the overall system performance, the average response time of the system is controlled between 12 and 18 ms, which can quickly adapt to the dynamic changes of different types of noise, ensuring that the system responds in real time and accurately cancels out noise. At the same time, the phase matching error of the system is controlled within the range of 2 to 4°. This indicator is at a leading level in active noise cancellation technology, further verifying the technical advantages of the present invention in high-precision acoustic phase control. In addition, through real-time optimization of the feedback control module, the system can dynamically adjust the power, phase and emission direction of the acoustic wave transmitter according to the noise residual value and performance indicators, thereby maximizing the noise reduction effect and system stability.

[0116] In summary, the present invention effectively solves the problems of limited noise coverage, low noise classification accuracy and insufficient real-time control capabilities in the prior art through technical means such as three-dimensional sound field modeling, deep learning noise classification and adaptive beam adjustment. In the experimental data of office scenes, whether it is steady-state noise, transient noise, periodic noise, or external complex noise, the system has achieved a noise reduction amplitude of 12 to 20 dB, and has fast response and high-precision phase matching capabilities. This series of data and performance fully demonstrates the effectiveness, adaptability and advancement of the present invention, which is suitable for a variety of noise control scenarios and provides users with a quiet, comfortable and stable acoustic environment.

[0117] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A method for adaptive depth adjustment of omnidirectional sound wave noise cancellation, characterized in that: The steps include: S1. Install acoustic wave transmitters at the four corners of the target space and multiple preset positions; S2, monitoring the noise level, acoustic characteristics and sound source distribution in the target space in real time through the acoustic sensor built into the sound wave transmitter, and generating acoustic environment data containing noise time domain, frequency domain and spatial information; S3. Based on the acoustic environment data, the spatial position, propagation path and energy distribution of the noise source are calculated using the time difference positioning algorithm and the beamforming algorithm to construct a three-dimensional sound field model; S4. Analyze the three-dimensional sound field model through the ResNet50 network model, identify the noise type, and generate a noise reduction strategy corresponding to the noise type; S5. Based on the noise cancellation strategy, an adaptive beam adjustment algorithm is used to control the vibration mode, phase and amplitude of each sound wave transmitter to generate a noise cancellation wave with a specific direction and shape, and to achieve phase cancellation between the noise cancellation wave and the target noise.

2. The method for adaptive depth adjustment of omnidirectional acoustic noise cancellation according to claim 1, characterized in that: The S3 specifically includes: S31. Use a time difference positioning algorithm to calculate the spatial position of the noise source, using the time difference of noise signals received by different microphones in the microphone array of the acoustic sensor: Among them, Δt ij It represents the time difference between microphone i and microphone j receiving the noise signal, t i and t j They represent the time when the corresponding microphone receives the signal, d i and d j Respectively represent the distances from the noise source to microphone i and microphone j, and c represents the speed of sound; S32, combining the time difference positioning result of the noise source with the spatial coordinate data, and using the geometric triangulation positioning method to calculate the specific position of the noise source in the three-dimensional space: Among them, (x s ,y s ,z s ) represents the three-dimensional spatial coordinates of the noise source, (x i ,y i ,z i ) represents the coordinate position of microphone i, t i represents the time it takes for the noise signal to propagate to microphone i, and N represents the number of microphones; S33. Based on the three-dimensional positioning result of the noise source, calculate the energy distribution of the noise source in the target space: Where P(f) represents power spectrum density, X(f,t) represents frequency domain data, and T represents time period; S34, modeling the propagation path of the noise signal through a beamforming algorithm to determine the direction of noise propagation and the main reflection path; S35. Construct a three-dimensional sound field model in the target space according to the spatial coordinates, propagation path and energy distribution of the noise source, wherein the three-dimensional sound field model includes the positioning coordinates of the noise source, the noise intensity distribution and the noise propagation path data.

3. The method for adaptive depth adjustment of omnidirectional acoustic noise cancellation according to claim 1, characterized in that: The S4 specifically includes: S41, inputting data into the three-dimensional sound field model to obtain the spatial position, propagation path, energy distribution and noise intensity data of the noise source; S42, extracting time domain features and spectrum features of the noise signal, wherein the time domain features include peak value, mean value and pulse amplitude; S43, inputting the extracted noise signal features into the ResNet50 network model, and classifying the noise types using the time domain features and spectrum features of the noise, wherein the noise types include steady-state noise, transient noise, and periodic noise; S44. Generate a noise reduction strategy corresponding to the noise type according to the classification result, wherein the noise reduction strategy includes a noise reduction target frequency band, a beam direction, and a vibrator parameter setting.

4. The method for adaptive depth adjustment of omnidirectional acoustic noise cancellation according to claim 1, characterized in that: The S5 specifically includes: S51, according to the noise cancellation strategy, obtain the type of noise, target frequency band, sound field distribution and noise propagation path, and determine the emission parameters of each sound wave transmitter, wherein the emission parameters include the vibration mode, amplitude, phase and emission direction of the vibrator; S52, controlling the vibrator arrays of multiple acoustic wave transmitters by an adaptive beam adjustment algorithm: Among them, y(t) represents the output signal of the acoustic wave transmitter, w i represents the control weight of the vibrator, x i (t) represents the input signal of the i-th vibrator, and N represents the number of vibrators; S53, adjusting the vibration amplitude and phase of each vibrator, and controlling the directivity of the sound wave by a beamforming method; Among them, w opt represents the optimal weight, d(t) represents the expected denoised signal, and E represents the mathematical expectation; S54, optimizing the emission direction of the beam based on the characteristics of noise propagation in the target space; Among them, P(θ) represents the acoustic wave transmission power, θ represents the transmission direction angle of the beam, and d i represents the distance from the ith vibrator to the reference point, and k represents the wave number; S55. Combined with the real-time propagation path and energy distribution data of noise, dynamically adjust the phase and amplitude of each sound wave transmitter to generate noise-cancelling waves in a specific direction and shape.

5. The method for adaptive depth adjustment of omnidirectional sound wave noise cancellation according to claim 1, characterized in that: The sound wave transmitter is composed of a plurality of small vibrators.

6. An omnidirectional sound wave noise cancellation system, implementing the omnidirectional sound wave noise cancellation adaptive depth adjustment method according to any one of claims 1 to 5, characterized in that: include: Multiple acoustic wave transmitters for transmitting noise-cancelling waves and collecting acoustic environment data; Data acquisition module, used to collect time domain, frequency domain and spatial information of noise; A data processing module is used to analyze and process the monitored data; A noise classification module, used to classify noise types; The adaptive beam adjustment module is used to adjust the vibration mode, phase and transmission power of the acoustic wave transmitter according to the noise cancellation strategy.

Images