Noise measuring device and noise filtering method thereof

Abstract

The invention discloses a noise measuring device and a noise filtering method thereof, and belongs to the technical field of acoustic signal processing. The noise measuring device comprises an annular microphone array, directional microphones and a processor; the annular microphone array and the directional microphone are electrically connected with the processor; the annular microphone array comprises a plurality of omnidirectional microphones arranged in an annular array, one part of the omnidirectional microphones correspond to the non-interested area, and the other part of the omnidirectional microphones correspond to the interested area; the directional microphone is arranged in the center of the annular microphone array, the directional microphone has a target sound receiving direction, and the target sound receiving direction corresponds to the region of interest; by adopting a plurality of omnidirectional microphones which are annularly arranged and a directional microphone, the device can comprehensively utilize the information of a plurality of microphones, and a beam forming algorithm is combined, so that a noise source in a factory boundary is accurately positioned. The method has flexibility and adaptability, and has a wide application prospect in the aspects of noise detection, voice recognition and audio monitoring.

Description

Technical Field

[0001] The invention belongs to the field of measurement technology, and particularly relates to a noise measurement device and a noise filtering method thereof. Background Art

[0002] In order to better solve and improve the problem of noise pollution, it is generally necessary to measure the noise at the factory boundary of industrial enterprises. However, the existing factory boundary noise equipment does have some technical challenges and limitations when measuring.

[0003] Most of the existing plant boundary noise equipment uses sound level meters for measurement. When measuring, it cannot distinguish whether the sound comes from inside or outside the plant boundary, and mistakenly counts the external noise into the noise level inside the plant boundary, resulting in inaccurate measurement. Since the sound level meter cannot distinguish the sound source, its measurement results may be interfered by external noise in complex sound environments, such as industrial areas, thus affecting the accuracy of the data. This inaccuracy may lead to misjudgment of the degree of noise pollution, which in turn affects subsequent noise control measures. In addition, although the sound level meter can measure the level of sound, it cannot provide detailed information about the intensity and frequency characteristics of the sound. Therefore, in a complex industrial environment, this device may not accurately capture all relevant noise parameters, thus affecting the measurement accuracy. The sound level meter mainly focuses on the measurement of sound levels, while ignoring other important characteristics of noise, such as the direction of sound propagation, the type of sound source, etc. This limitation restricts the comprehensive analysis and solution of noise pollution problems.

[0004] In view of the technical problem that the existing sound level meters cannot distinguish whether the sound comes from inside or outside the factory boundary when measuring factory boundary noise, resulting in inaccurate measurement, it is urgent to find a noise measurement device and its noise filtering method to improve the accuracy and reliability of sound source positioning. Summary of the invention

[0005] In order to overcome the shortcomings of the above-mentioned prior art, the purpose of the present invention is to provide a noise measuring device and a noise filtering method thereof, so as to solve the technical problem that the existing sound level meter cannot distinguish whether the sound comes from inside or outside the factory boundary when measuring the factory boundary noise, resulting in inaccurate measurement.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] The invention discloses a noise measurement device, comprising a ring microphone array, a directional microphone and a processor;

[0008] The annular microphone array includes a plurality of omnidirectional microphones arranged in an annular array, wherein a portion of the omnidirectional microphones corresponds to a non-interest area, and another portion of the omnidirectional microphones corresponds to an interest area;

[0009] The directional microphone is arranged at the center of the circular microphone array, and the directional microphone has a target sound receiving direction, and the target sound receiving direction corresponds to the region of interest;

[0010] The annular microphone array is used to collect first sound source information;

[0011] The directional microphone is used to collect the second sound source information;

[0012] The processor is used to calculate the first sound source information to obtain the specific sound source information corresponding to the region of interest; and determine the sound information of the sound source in the region of interest based on the specific sound source information and the second sound source information.

[0013] Preferably, it further comprises a substrate, a plurality of omnidirectional microphones are arranged on the substrate in a circular array, a directional microphone is arranged at the center of the substrate, the sound receiving direction of the directional microphone is arranged in the radial direction of the circular array, and a marking area corresponding to the sound receiving direction is arranged on the substrate;

[0014] The multiple omnidirectional microphones are all located in a specifically determined positional relationship relative to the identification area.

[0015] Preferably, the distance between any two adjacent omnidirectional microphones in the annular microphone array is between 1 / 2 and 1 times the wavelength of the main sound source in the region of interest.

[0016] The present invention also discloses a noise filtering method, which uses the above-mentioned noise measurement device to filter noise, comprising the following steps:

[0017] S1, collecting first sound source information through a circular microphone array, and collecting second sound source information through a directional microphone;

[0018] S2, the processor pre-processes the first sound source information and the second sound source information collected in step S1, and aligns the first sound source information and the second sound source information on the time axis;

[0019] S3, the processor calculates the first sound source information according to the beamforming algorithm to obtain specific sound source information corresponding to the region of interest;

[0020] S4. The processor determines the sound information of the sound source in the area of ​​interest based on the specific sound source information obtained in step S3 and the second sound source information preprocessed in step S2, and completes noise filtering.

[0021] Preferably, in step S1, the first sound source information is multi-channel data, including the sound magnitude and phase of the sound source in the area of ​​interest, and the sound magnitude and phase of the sound source in the area of ​​no interest.

[0022] Preferably, in step S3, according to a beamforming algorithm, a processor calculates the first sound source information to obtain specific sound source information corresponding to the region of interest, which specifically includes the following steps:

[0023] S31, determining the azimuth range of the region of interest;

[0024] S32, calculating a weight vector pointing to the region of interest through a covariance matrix according to the multi-channel data in the first sound source information;

[0025] S33, performing a dot multiplication on the multi-channel data in the first sound source information and the weight vector pointing to the region of interest calculated in step S32, and adding the results to obtain a synthetic signal pointing to the region of interest;

[0026] S34, pre-processing the synthetic signal pointing to the region of interest obtained in step S33 to obtain specific sound source information corresponding to the region of interest.

[0027] Further preferably, in step S4, the processor fuses the specific sound source information obtained in step S3 and the second sound source information preprocessed in step S2 by weighted fusion to obtain the sound information of the sound source in the area of ​​interest, thereby completing noise filtering;

[0028] The specific steps include:

[0029] The specific sound source information and the second sound source information preprocessed in step S2 are weightedly fused using the following formula:

[0030] W=A*W1+B*W2

[0031] Wherein, W represents the size of the sound in the sound information of the sound source in the area of ​​interest; W1 represents the size of the sound in the specific sound source information; W2 represents the size of the sound in the second sound source information; A and B are weighting coefficients, and the value range is 0≤A≤1, 0≤B≤1, and A+B=1.

[0032] Further preferably, the values ​​of A1 and B are determined according to the following steps:

[0033] First, determine the frequency of the second sound source information; then determine the values ​​of A and B based on the frequency of the second sound source information.

[0034] Further preferably, the step of determining the values ​​of A and B according to the frequency of the second sound source information includes:

[0035] Determining whether the frequency of the second sound source information is greater than a preset value;

[0036] If the frequency of the second sound source information is greater than a preset value, it is determined that the second sound source is a high frequency, and B satisfies 0.8≤B≤1.

[0037] If the frequency of the second sound source information is less than a preset value, it is determined that the second sound source is a low frequency, and the value of B is B≤0.5.

[0038] Further preferably, the preset value is 200 Hz.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] The invention discloses a noise measurement device, comprising: a circular microphone array, a directional microphone and a processor; the circular microphone array and the directional microphone are both electrically connected to the processor; the circular microphone array comprises a plurality of omnidirectional microphones arranged in a circular array, which can capture sound in 360 degrees. Among them, a part of the omnidirectional microphones corresponds to a non-interested area, and is used to collect environmental noise or other non-target sounds for subsequent noise elimination; another part of the omnidirectional microphones corresponds to an area of ​​interest; it focuses on capturing sound signals in the area, which is convenient for subsequent sound source positioning and sound analysis; the directional microphone is arranged at the center of the circular microphone array, and the directional microphone has a target sound receiving direction, which corresponds to the area of ​​interest; it can more accurately pick up sound from a specific direction and reduce interference from other directions. Several circular microphone arrays are distributed in a ring shape, and the central directional microphone is installed at the ring center of several circular microphone arrays. Combining the circular array directional microphone and the directional microphone, the accuracy and reliability of the sound source direction measurement can be significantly improved. By adopting several omnidirectional microphones arranged in a ring shape and a directional microphone, the device can comprehensively utilize the information of multiple microphones, combine the beamforming algorithm, and accurately locate the noise source within the factory boundary. Omnidirectional microphones can capture sound signals from all directions, while directional microphones can enhance the ability to capture sound signals in a specific direction. The combination of the two greatly improves the accuracy of sound source positioning. The noise measurement device disclosed in the present invention provides an efficient and accurate solution for the measurement and control of factory boundary noise through the advantages of high-precision sound source positioning, multi-directional coverage and analysis, stability and reliability, flexibility and adaptability, and easy installation and use.

[0041] Furthermore, it also includes a substrate, a plurality of omnidirectional microphones are arranged in a circular array on the substrate, a directional microphone is arranged at the center of the substrate, the sound receiving direction of the directional microphone is arranged in the radial direction of the circular array, and a marking area corresponding to the sound receiving direction is arranged on the substrate; the plurality of omnidirectional microphones are all in a specifically determined positional relationship relative to the marking area. During measurement, the substrate is arranged horizontally so that each omnidirectional microphone corresponds to a different orientation in the plane, and the marking area corresponds to the area of ​​interest at the factory boundary.

[0042] Furthermore, since several ring microphones are arranged at equal intervals, they can correspond to the four directions of up, down, left, and right, thus achieving full coverage of sound signals around the factory boundary. This layout enables the device to capture noise from multiple directions at the same time, providing comprehensive data support for subsequent noise analysis and control.

[0043] Furthermore, the selected microphones have highly consistent sensitivity, frequency response and phase characteristics, ensuring the stability and reliability of the measurement results. In addition, through appropriate wiring methods and signal processing units, low noise, low crosstalk and high signal-to-noise ratio are guaranteed during signal transmission, further improving the measurement accuracy and reliability of the device.

[0044] Furthermore, the layout and configuration of the device have a certain degree of flexibility and can be adjusted and optimized according to actual needs. For example, the number and spacing of microphones can be adjusted according to the specific shape and size of the factory boundary to adapt to different measurement environments. At the same time, the device is also suitable for a variety of noise measurement and control scenarios, with strong versatility and adaptability.

[0045] Furthermore, the vertical installation method can be conveniently installed within 5 meters of the factory boundary, and the directional microphone is facing the noise-sensitive area of ​​the factory boundary. This installation method is not only easy to operate and maintain, but also ensures that the device can effectively capture the noise signal within the factory boundary.

[0046] Furthermore, by optimizing the configuration of the circular array, including increasing the number of microphones and adjusting the spacing and layout, the resolution and accuracy of sound source direction estimation are improved, which helps to more accurately identify specific sound sources and reduce confusion about adjacent or nearby sound sources.

[0047] The present invention also discloses a noise filtering method, which can accurately determine the direction of the sound source by combining the information of the directional microphone and the annular microphone array, thereby improving the accuracy of the sound source positioning. This helps to more accurately identify the noise source within the factory boundary, and provides strong support for the subsequent noise control. The denoising and gain adjustment operations in the preprocessing step effectively improve the signal-to-noise ratio of the useful signal, making the subsequent signal processing and analysis more accurate and reliable. At the same time, the synchronous operation ensures the alignment of all microphone signals on the time axis, avoiding signal distortion caused by time dislocation. Through the spatial filtering processing technology, the noise outside the factory boundary can be accurately separated, so as to obtain the accurate noise within the factory boundary. This helps to better understand the noise conditions within the factory boundary and provides a basis for formulating targeted noise control strategies. Through accurate noise positioning and separation, it can not only be used for the measurement and control of factory boundary noise, but also can be widely used in other places with serious noise pollution, such as factories, airports, and major traffic arteries. At the same time, its results can also be used in multiple fields such as noise source identification, noise source tracking, and noise control strategy formulation, providing strong support for environmental protection and noise control. The noise filtering method disclosed in the present invention provides an efficient and accurate method for measuring and controlling factory boundary noise by improving the accuracy of sound source positioning, enhancing signal quality, achieving noise separation and target tracking, and having advantages such as flexibility and adaptability and multi-field application. It improves the performance of the sound source positioning system, making it have a wider application prospect in the fields of noise detection, speech recognition, audio monitoring, etc.

[0048] Furthermore, by adopting advanced signal processing algorithms, the interference of non-direct sound components such as background noise, reflected sound, and reverberation on the estimation of the sound source direction is effectively reduced, which helps to more accurately extract and locate the target sound source in a complex sound field environment.

[0049] Furthermore, for sound sources with extended and non-uniform radiation characteristics, by optimizing the array design and signal processing algorithm, its complete spatial characteristics can be better captured, thereby improving the accuracy of DOA estimation. Improving existing signal processing algorithms or introducing new algorithms can reduce sensitivity to initial conditions, enhance noise suppression capabilities, and reduce assumptions about the number of sound sources. This helps to achieve accurate sound source direction estimation in a wider range of scenarios.

[0050] Furthermore, by introducing directional microphones or other reference devices, additional information can be provided for correcting hardware errors in the circular array, such as nonlinear response, sensitivity inconsistency, and phase mismatch, which helps to reduce the impact of hardware non-ideal characteristics on the estimation of the sound source direction.

[0051] Furthermore, by optimizing the array design and signal processing algorithm, rapid response and accurate tracking of dynamic sound fields can be achieved. This helps to maintain a high level of sound source localization accuracy when the sound source or array environment changes. The information of the circular array and directional microphone is fused in a linear weighted manner, making the weight selection flexible. According to factors such as the trust in the two types of information and environmental conditions in actual applications, the weight coefficient can be adjusted to adapt to different noise environments and needs.

[0052] Furthermore, real-time tracking and power spectrum estimation of the target noise signal enable us to accurately grasp the size and changing trend of the noise, providing important data support for noise control. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a structural schematic diagram of a noise measurement device disclosed in the present invention;

[0054] Figure 2 A schematic diagram of a use scenario of the noise measurement device disclosed in the present invention;

[0055] Figure 3 It is a schematic diagram of the three-dimensional structure of the noise measurement device disclosed in the present invention;

[0056] Figure 4 A top view of the noise measuring device disclosed in the present invention;

[0057] Among them, 1- ring microphone array; 2- directional microphone. DETAILED DESCRIPTION

[0058] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0059] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0060] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0061] See also Figure 1 It is a structural schematic diagram of a noise measuring device disclosed in the present invention. As can be seen from the figure, the noise measuring device includes a ring-shaped microphone array 1 composed of four omnidirectional microphones and a directional microphone 2 located in the center; the ring-shaped microphone array 1 composed of four omnidirectional microphones is distributed in a ring shape and is arranged at equal intervals, and is respectively distributed above, below, on the left and on the right of the directional microphone 2, and the directional microphone 2 is installed at the ring center of the four omnidirectional microphones.

[0062] See also Figure 2 It is a schematic diagram of the use scenario of the noise measurement device disclosed in the present invention; as can be seen from the figure, the inside of the factory boundary is the area of ​​interest, the outside of the factory boundary is the area of ​​non-interest, and the noise measurement device is placed in the area of ​​interest.

[0063] See also Figure 3 Schematic diagram of the three-dimensional structure of the noise measurement device disclosed in the present invention; see Figure 4 This is a top view of the noise measurement device disclosed in the present invention; as can be seen from the figure, the overall design of the noise measurement device is that the central part of the device is a directional microphone, which has high sensitivity and directionality and can accurately capture sounds in a specific direction. In the recognition area of ​​the directional microphone, the directional microphone can most effectively receive the direction and angle of the sound. The recognition area of ​​the directional microphone is marked in a fan shape, which clearly shows the main direction in which the microphone receives sound. In this area, the microphone can capture the clearest and most accurate sound signal, which is crucial for noise measurement.

[0064] Example 1

[0065] A noise measurement device, comprising: a circular microphone array 1, a directional microphone 2 and a processor;

[0066] The annular microphone array 1 includes a plurality of omnidirectional microphones arranged in an annular array, wherein a portion of the omnidirectional microphones corresponds to a non-interest area, and another portion of the omnidirectional microphones corresponds to an interest area;

[0067] The directional microphone 2 is arranged at the center of the circular microphone array 1, and the directional microphone 2 has a target sound receiving direction, and the target sound receiving direction corresponds to the region of interest;

[0068] The annular microphone array 1 is used to collect first sound source information;

[0069] The directional microphone 2 is used to collect the second sound source information;

[0070] The processor is used to calculate the first sound source information to obtain the specific sound source information corresponding to the region of interest; and determine the sound information of the sound source in the region of interest based on the specific sound source information and the second sound source information.

[0071] The method for filtering noise using the noise measuring device comprises the following steps:

[0072] S1, collecting first sound source information through the circular microphone array 1, and collecting second sound source information through the directional microphone 2;

[0073] S2, the processor pre-processes the first sound source information and the second sound source information collected in step S1, and aligns the first sound source information and the second sound source information on the time axis;

[0074] S3, the processor calculates the first sound source information according to the beamforming algorithm to obtain specific sound source information corresponding to the region of interest;

[0075] S4. The processor determines the sound information of the sound source in the area of ​​interest based on the specific sound source information obtained in step S3 and the second sound source information preprocessed in step S2, and completes noise filtering.

[0076] Example 2

[0077] A noise measurement device, comprising: a circular microphone array 1, a directional microphone 2 and a processor;

[0078] The annular microphone array 1 includes a plurality of omnidirectional microphones arranged in an annular array, wherein a portion of the omnidirectional microphones corresponds to a non-interest area, and another portion of the omnidirectional microphones corresponds to an interest area;

[0079] The directional microphone 2 is arranged at the center of the circular microphone array 1, and the directional microphone 2 has a target sound receiving direction, and the target sound receiving direction corresponds to the region of interest;

[0080] The annular microphone array 1 is used to collect first sound source information;

[0081] The directional microphone 2 is used to collect the second sound source information;

[0082] The processor is used to calculate the first sound source information to obtain the specific sound source information corresponding to the region of interest; and determine the sound information of the sound source in the region of interest based on the specific sound source information and the second sound source information.

[0083] The directional microphone 2 is arranged at the center of the substrate, the sound receiving direction of the directional microphone 2 is arranged in the radial direction of the annular array, and a marking area corresponding to the sound receiving direction is arranged on the substrate;

[0084] The multiple omnidirectional microphones are all located in a specific positional relationship relative to the identification area;

[0085] The distance between any two adjacent omnidirectional microphones in the annular microphone array 1 is 1 times the wavelength of the main sound source in the region of interest.

[0086] The method for filtering noise using the noise measuring device comprises the following steps:

[0087] S1. Collect first sound source information through the circular microphone array 1, and collect second sound source information through the directional microphone 2; wherein the first sound source information is multi-channel data, including the sound size and phase of the sound source in the area of ​​interest, and the sound source size and phase in the area of ​​non-interest.

[0088] S2, the processor pre-processes the first sound source information and the second sound source information collected in step S1, and aligns the first sound source information and the second sound source information on the time axis;

[0089] S3, the processor calculates the first sound source information according to the beamforming algorithm to obtain specific sound source information corresponding to the region of interest; specifically comprising the following steps:

[0090] S31, determining the azimuth range of the region of interest;

[0091] S32, calculating a weight vector pointing to the region of interest through a covariance matrix according to the multi-channel data in the first sound source information;

[0092] S33, performing a dot multiplication on the multi-channel data in the first sound source information and the weight vector pointing to the region of interest calculated in step S32, and adding the results to obtain a synthetic signal pointing to the region of interest;

[0093] S34, pre-processing the synthetic signal pointing to the region of interest obtained in step S33 to obtain specific sound source information corresponding to the region of interest.

[0094] S4, the processor determines the sound information of the sound source in the area of ​​interest according to the specific sound source information obtained in step S3 and the second sound source information preprocessed in step S2, and completes noise filtering; comprising the following steps:

[0095] The specific sound source information and the second sound source information preprocessed in step S2 are weightedly fused using the following formula:

[0096] W=A*W1+B*W2

[0097] Wherein, W represents the size of the sound in the sound information of the sound source in the area of ​​interest; W1 represents the size of the sound in the specific sound source information; W2 represents the size of the sound in the second sound source information; A and B are weighting coefficients, and the value range is 0≤A≤1, 0≤B≤1, and A+B=1.

[0098] First determine the frequency of the second sound source information; then determine the values ​​of A and B based on the frequency of the second sound source information. Determine whether the frequency of the second sound source information is greater than a preset value; if the frequency of the second sound source information is greater than 200 Hz, determine that the second sound source is a high frequency, and B satisfies 0.8≤B≤1. If the frequency of the second sound source information is less than 200 Hz, determine that the second sound source is a low frequency, and the value of B is B≤0.5.

[0099] Example 3

[0100] A noise measurement device comprises: a circular microphone array 1, a directional microphone 2 and a processor; the circular microphone array 1 and the directional microphone 2 are both electrically connected to the processor; the circular microphone array 1 comprises a plurality of omnidirectional microphones arranged in a circular array, wherein a part of the omnidirectional microphones corresponds to a non-interested area, and another part of the omnidirectional microphones corresponds to a region of interest; the directional microphone 2 is arranged at the center of the circular microphone array 1, and the directional microphone 2 has a target sound receiving direction, and the target sound receiving direction corresponds to the region of interest; the processor processes the first sound source information collected by the circular microphone array 1 according to a beamforming algorithm to obtain specific sound source information corresponding to the region of interest; wherein the first sound source information comprises the sound source information of the region of interest and the sound source information of the non-interested area; the processor determines the sound level of the region of interest according to the specific sound source information and the second sound source information collected by the directional microphone 2. The noise measurement device also includes a substrate, a plurality of omnidirectional microphones are arranged in a circular array on the substrate, a directional microphone 2 is arranged on the substrate, the sound receiving direction of the directional microphone 2 is toward the radial direction of the circular array, and the substrate is provided with a marking area corresponding to the sound receiving direction; the plurality of omnidirectional microphones are all in a specific positional relationship relative to the marking area; during measurement, the substrate is arranged horizontally so that each omnidirectional microphone corresponds to a different orientation in the plane, and the marking area corresponds to the area of ​​interest at the factory boundary. The spacing between any two adjacent omnidirectional microphones in the circular microphone array 1 is configured to be between 1 / 2 and 1 times the wavelength of the main sound source in the area of ​​interest.

[0101] The method for filtering noise using the noise measuring device comprises the following steps:

[0102] S1, obtaining first sound source information collected by the annular microphone array 1 and obtaining second sound source information collected by the directional microphone 2; wherein the first sound source information includes the sound size and phase of the sound source in the area of ​​interest and the sound source size and phase in the area of ​​non-interest, which is multi-channel data;

[0103] S2, preprocessing the first sound source information and the second sound source information, and aligning the first sound source information and the second sound source information on the time axis;

[0104] S3, calculating the first sound source information according to a beamforming algorithm to obtain specific sound source information corresponding to the region of interest;

[0105] S4. Determine the sound information of the sound source in the area of ​​interest based on the specific sound source information and the second sound source information.

[0106] Wherein, step S3 comprises the following steps:

[0107] S31, determining the azimuth range of the region of interest;

[0108] S32, calculating a weight vector pointing to the region of interest through a covariance matrix according to the multi-channel signal in the first sound source information;

[0109] S33, performing dot multiplication on the information of each channel in the first sound source information and the calculated weight vector and adding the results to obtain a synthetic signal pointing to the region of interest;

[0110] S34, pre-processing the synthesized signal to obtain specific sound source information in the region of interest.

[0111] Wherein, in step S4, determining the sound information of the sound source of the region of interest based on the specific sound source information and the second sound source information includes the following steps: fusing the specific sound source information and the second sound source information by weighted fusion to obtain the sound information of the sound source of the region of interest.

[0112] The arrangement of the sound information of the sound source in the region of interest includes weighted fusion of the specific sound source information and the second sound source information using the following formula:

[0113] W=A*W1+B*W2

[0114] Wherein, W represents the size of the sound in the sound information of the sound source in the area of ​​interest; W1 represents the size of the sound in the specific sound source information; W2 represents the size of the sound in the second sound source information; A and B are weighting coefficients, and the value range is 0≤A≤1, 0≤B≤1, and A+B=1.

[0115] In the formula, the values ​​of A1 and B are determined according to the following steps: first determine the frequency of the second sound source information, and then determine the values ​​of A and B according to the frequency of the second sound source information.

[0116] The step of determining the values ​​of A and B according to the frequency of the second sound source information includes:

[0117] First, determine whether the frequency of the second sound source information is greater than a preset value;

[0118] If the frequency of the second sound source information is greater than the preset value, the second sound source is determined to be high frequency, and B satisfies 0.8≤B≤1;

[0119] If the frequency of the second sound source information is less than a preset value, it is determined that the second sound source is a low frequency, and the value of B is B≤0.5; wherein the preset value is 200 Hz.

[0120] Example 4

[0121] A noise measurement device includes: 4 omnidirectional microphones arranged in a ring shape and a directional microphone 2 arranged in the center of the ring. The 4 ring microphones are arranged on a circular circuit board. They are arranged at equal intervals, and each microphone corresponds to a filter circuit. The directional microphone 2 is arranged in the center of the circuit board. When there are 4 ring microphones, they can correspond to four directions of up, down, left and right. In this way, the noise within the factory boundary can be accurately obtained according to the direction of the sound source according to the beamforming algorithm and the directional sound source information obtained by the directional microphone 2.

[0122] The circular array design is adopted, and a directional microphone 2 is set in the middle of the array. The microphone selection and layout of the entire system ensure the stability and accuracy of the array performance. The selected series of microphones need to have highly consistent sensitivity, frequency response and phase characteristics. The microphones in the array are omnidirectional microphones, which can better capture sound signals from all directions for subsequent beamforming and other processing.

[0123] Determine the spacing between microphones based on the frequency range of the sound signal of interest and the desired resolution. For low and mid-frequency bands, the spacing is between 1 / 2 and 1 times the wavelength of the sound wave; high frequency bands may require a smaller spacing to improve resolution.

[0124] The central directional microphone 2 is a supercardioid or cardioid microphone with strong directivity to enhance the ability to capture sound signals in a specific direction. The characteristics of the central directional microphone 2 should be well matched with the omnidirectional microphone in the circular microphone array 1 to avoid excessive differences that cause overall response distortion.

[0125] The circular microphone array 1 and the central directional microphone 2 are connected to the signal processing unit (such as FPGA, DSP) through appropriate wiring, and low noise, low crosstalk and high signal-to-noise ratio are guaranteed during signal transmission. When in use, it is installed vertically within 5 meters of the factory boundary so that the directional microphone 2 faces the noise-sensitive area of ​​the factory boundary.

[0126] The method for filtering noise using the noise measuring device comprises the following steps:

[0127] The sound source is pre-positioned using the circular microphone array 1, and the signal gain in the target direction is optimized in combination with the data from the central directional microphone 2, ultimately obtaining the noise within the factory boundary.

[0128] The central directional microphone 2 and the circular microphone array 1 are used to identify the direction of the noise. The following steps are performed:

[0129] 1) Signal acquisition: The central directional microphone 2 and the omnidirectional microphones in the circular microphone array 1 simultaneously collect sound signals and convert the analog audio signals into digital signals through an analog / digital converter (ADC).

[0130] 2) Preprocessing: Preprocess the digital signals collected by the central directional microphone 2 and the circular microphone array 1, including denoising, gain adjustment, and synchronization, to ensure that all microphone signals are aligned on the time axis.

[0131] Among them, denoising can improve the signal-to-noise ratio of useful signals, and gain adjustment can keep the signal levels of each channel as consistent as possible. Ensure that the sampling rates of the central directional microphone 2 and the circular microphone array 1 are consistent, otherwise it will cause misalignment on the time axis and affect the correct superposition and processing of the signals. Due to factors such as hardware clock drift and signal transmission delay, the time when the central directional microphone 2 and the circular microphone array 1 signals arrive at the processing system may be slightly different. Accurate timestamp synchronization technology is required to ensure that all signals are accurately aligned on the time axis.

[0132] 3) Applying the beamforming algorithm, based on the signal strength and phase information received by the central directional microphone 2 and the circular microphone array 1, the signal of the sound source pointing in a specific direction is calculated.

[0133] Specifically, the covariance matrix of the array is calculated. The covariance matrix contains the correlation information of each channel signal and reflects the arrival direction of the sound source. The MUSIC spatial spectrum estimation method is applied (the eigenvalue decomposition of the covariance matrix is ​​calculated to separate the signal subspace and the noise subspace. The MUSIC spectrum function is constructed, which presents a peak in the direction of the sound source and approaches zero in the direction of the noise subspace. The maximum value or peak of the MUSIC spectrum is found. The azimuth angle corresponding to these peaks is the DOA estimate of the sound source.) The DOA information of the sound source is extracted from the covariance matrix. According to the DOA information, the weighting vector (beamforming vector) is calculated to perform weighted processing on the signals of each channel. The weighting vector is usually calculated using the MVDR beamforming algorithm. The weighting vector is used to linearly combine the original microphone signals to form a "beam" pointing to a specific sound source direction, thereby enhancing the signal within the factory boundary and suppressing interference in other directions.

[0134] 4) Refine the sound source direction in combination with the data provided by directional microphone 2.

[0135] The directional microphone 2 can clearly determine the direction of the sound and obtain the source of the main noise.

[0136] Ring microphone array 1 monitors the sound from different angles and accurately determines the angle of the noise through complex mathematical calculations, namely DOA estimation. The information of directional microphone 2 and ring microphone array 1 is combined to find the source of the noise more accurately. Ring microphone array 1 provides general direction clues, and directional microphone 2 provides more precise location clues. Specifically, different weights are assigned to the information of directional microphone 2 and ring microphone array 1, and then the weighted information is added together. For example, if directional microphone 2 is very sure that the noise is coming from the southeast, it will be given a higher weight; ring microphone array 1 calculates a more precise angle of the noise and also gives it a weight. Then, according to their weight ratio, the two pieces of information are added together to get the most likely source direction of the noise. This is the process of combining the DOA estimation of directional microphone 2 and ring microphone array 1 to improve the accuracy of sound source localization by weighted fusion of the two types of information. This not only takes advantage of the keen perception of the general direction of directional microphone 2, but also takes advantage of the precise calculation of the precise angle by ring microphone array 1, making noise localization more accurate.

[0137] Before fusion, the signal obtained by directional microphone 2 usually needs to be preprocessed to ensure that it can be effectively fused with the DOA estimation result of circular microphone array 1. Preprocessing includes: 1. Noise suppression: remove non-target signal components such as environmental noise, power supply noise, electromagnetic interference, etc. in the signal of directional microphone 2, improve the signal-to-noise ratio, and ensure the purity of the signal. 2. Directivity calibration: ensure that the directivity of directional microphone 2 is consistent with the actual installation angle, without offset or rotation error. If necessary, calibrate to ensure its directivity accuracy. 3. Signal strength evaluation: measure the signal strength received by directional microphone 2 to determine whether it meets the "strong" standard. This helps to determine whether this signal should be included in the fusion process. 4. Feature extraction: extract features related to the direction of the sound source from the signal of directional microphone 2, such as phase difference, amplitude ratio, etc., so as to compare and fuse with the array DOA estimation result. 5. Standardization or normalization: If the signal dimensions, units or dynamic ranges used by directional microphone 2 and circular microphone array 1 are different, standardization or normalization may be required to make the two comparable. 6. Time domain or frequency domain alignment: Ensure that the directional microphone 2 signal is aligned with the circular microphone array 1 signal on the time axis, or have the same analysis window in the frequency domain, so as to facilitate effective fusion. Through the above preprocessing steps, it can be ensured that the signal obtained by the directional microphone 2 has a good signal-to-noise ratio, accurate directionality, reasonable strength evaluation, comparable features and consistency with the circular microphone array 1 signal before fusion, which is conducive to improving the accuracy and reliability of the fusion result.

[0138] If directional microphone 2 captures a strong sound signal and its directional characteristics indicate a certain direction, it can be considered that the direction is the main sound source. (It means that there is noise in the factory boundary.) Then, the DOA estimation result of the circular microphone array 1 is combined with the directional characteristics of the directional microphone 2, and the positioning accuracy can be further improved by weighted fusion of the two types of information.

[0139] The two types of information are fused in a linear weighted manner: the DOA estimation result of the circular microphone array 1 and the signal measured by the directional microphone 2 are weighted and fused, which means that the information from these two sources is integrated according to a certain weight ratio. The choice of weight is usually based on the evaluation of the accuracy and reliability of the two information sources, as well as the needs of practical applications. The fused sound source direction estimation value = circular microphone array DOA weight × circular microphone array DOA estimation value + directional microphone weight × direction indicated by the directional microphone. The sum of the two weights should be equal to 1 to ensure the rationality of the fusion result. The choice of weight coefficient can be adjusted according to factors such as the trust in the two types of information and environmental conditions in practical applications. For example, if the DOA estimation accuracy of the circular microphone array 1 is high and stable, a larger weight can be given.

[0140] The fusion result is a new and more accurate estimate of the sound source direction, which comprehensively considers the multi-angle information provided by the circular microphone array 1 and the directional information provided by the directional microphone 2, thereby improving the accuracy and reliability of sound source localization. This result can be used for subsequent noise source identification, noise source tracking, noise control strategy formulation, etc.

[0141] 5) Noise separation and target tracking: Based on the identified noise direction, spatial filtering processing technology is used to extract the target noise signal from the overall noise to determine the target noise size.

[0142] According to the identified noise direction, the sound field is divided into two parts, the target noise signal and the background noise, using a spatial filter. The separated target noise signal is tracked in real time. During the tracking process, the power spectrum of the target noise signal is estimated or the energy is calculated to determine the size of the target noise. This can be done by performing spectrum analysis such as short-time Fourier transform (STFT) and Welch method to calculate the energy of the noise signal in different frequencies or time windows, thereby obtaining an estimate of the noise size.

[0143] The above method can accurately separate the noise outside the factory boundary, thereby obtaining the accurate noise within the factory boundary.

[0144] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A noise measuring device, characterized in that: It comprises a circular microphone array (1), a directional microphone (2) and a processor; The annular microphone array (1) comprises a plurality of omnidirectional microphones arranged in an annular array, wherein a portion of the omnidirectional microphones corresponds to a non-interest area, and another portion of the omnidirectional microphones corresponds to an interest area; The directional microphone (2) is arranged at the center of the annular microphone array (1), and the directional microphone (2) has a target sound collection direction, and the target sound collection direction corresponds to the area of ​​interest; The annular microphone array (1) is used to collect first sound source information; The directional microphone (2) is used to collect second sound source information; The processor is used to calculate the first sound source information to obtain the specific sound source information corresponding to the region of interest; and determine the sound information of the sound source in the region of interest according to the specific sound source information and the second sound source information.

2. The noise measuring device according to claim 1, characterized in that: It also comprises a substrate, the plurality of omnidirectional microphones are arranged on the substrate in a circular array, the directional microphone (2) is arranged at the center of the substrate, the sound collection direction of the directional microphone (2) is arranged in the radial direction of the circular array, and a marking area corresponding to the sound collection direction is arranged on the substrate; The multiple omnidirectional microphones are all in a specifically determined positional relationship relative to the identification area.

3. The noise measuring device according to claim 1, characterized in that: The distance between any two adjacent omnidirectional microphones in the annular microphone array (1) is between 1 / 2 and 1 times the wavelength of the main sound source in the area of ​​interest.

4. A noise filtering method, characterized in that: The noise filtering is performed using the noise measurement device according to any one of claims 1 to 3, comprising the following steps: S1, collecting first sound source information through a circular microphone array (1), and collecting second sound source information through a directional microphone (2); S2, the processor pre-processes the first sound source information and the second sound source information collected in step S1, and aligns the first sound source information and the second sound source information on the time axis; S3, the processor calculates the first sound source information according to the beamforming algorithm to obtain specific sound source information corresponding to the region of interest; S4. The processor determines the sound information of the sound source in the area of ​​interest based on the specific sound source information obtained in step S3 and the second sound source information preprocessed in step S2, and completes noise filtering.

5. The noise filtering method according to claim 4, characterized in that: In step S1, the first sound source information is multi-channel data, including the sound size and phase of the sound source in the area of ​​interest, and the sound size and phase of the sound source in the area of ​​no interest.

6. The noise filtering method according to claim 4, characterized in that: In step S3, according to the beamforming algorithm, the processor calculates the first sound source information to obtain specific sound source information corresponding to the region of interest, which specifically includes the following steps: S31, determining the azimuth range of the region of interest; S32, calculating a weight vector pointing to the region of interest through a covariance matrix according to the multi-channel data in the first sound source information; S33, performing a dot multiplication on the multi-channel data in the first sound source information and the weight vector pointing to the region of interest calculated in step S32, and adding the results to obtain a synthetic signal pointing to the region of interest; S34, pre-processing the synthetic signal pointing to the region of interest obtained in step S33 to obtain specific sound source information corresponding to the region of interest.

7. The noise filtering method according to claim 6, characterized in that: In step S4, the processor fuses the specific sound source information obtained in step S3 and the second sound source information preprocessed in step S2 by weighted fusion to obtain the sound information of the sound source in the region of interest, thereby completing noise filtering; The specific steps include: The specific sound source information and the second sound source information preprocessed in step S2 are weightedly fused using the following formula: W=A*W1+B*W2 Wherein, W represents the size of the sound in the sound information of the sound source in the area of ​​interest; W1 represents the size of the sound in the specific sound source information; W2 represents the size of the sound in the second sound source information; A and B are weighting coefficients, and the value range is 0≤A≤1, 0≤B≤1, and A+B=1.

8. The noise filtering method according to claim 7, characterized in that: The values ​​of A1 and B are determined according to the following steps: First, determine the frequency of the second sound source information; then determine the values ​​of A and B based on the frequency of the second sound source information.

9. The noise filtering method according to claim 8, characterized in that: The step of determining the values ​​of A and B according to the frequency of the second sound source information includes: Determining whether the frequency of the second sound source information is greater than a preset value; If the frequency of the second sound source information is greater than a preset value, it is determined that the second sound source is a high frequency, and B satisfies 0.8≤B≤1. If the frequency of the second sound source information is less than a preset value, it is determined that the second sound source is a low frequency, and the value of B is B≤0.

5.

10. The noise filtering method according to claim 9, characterized in that: The preset value is 200 Hz.

Images