Do you know any way to locate the noise source quickly and accurately?

There are many methods to identify noise sources, which are quite different in complexity, accuracy and cost. The actual use time can be determined according to the specific requirements of the research object and the possible conditions of manpower and material resources.

Specifically, noise source identification methods can be roughly divided into two categories:?

The first category is conventional acoustic measurement and analysis methods, including separation operation method, separation covering method, near-field measurement method, surface velocity measurement method and so on.

The second category is the acoustic signal processing method developed based on modern signal analysis theory. Sound intensity method, area intensity method, spectrum analysis, cepstrum analysis, cross-correlation and cross-spectrum analysis, coherence analysis and so on all belong to this category. ?

According to the complexity of sound source and the requirements of research work, different identification methods can be selected or used in combination at different research stages. Acoustic measurement The human auditory system can distinguish different sounds more accurately than the most complex noise measurement system. After long-term practice, it is possible for people to subjectively judge the frequency and location of noise sources. Experienced operators and inspectors can judge whether it is normal or not from the noise of the machine running on the production site, and can judge the cause of the abnormality. This subjective evaluation method is often very useful in production practice. In order to avoid other interference factors, medical stethoscope can also be used. But not everyone can reach the result of subjective judgment, because it has subjective factors, the same machine noise, and the recognition results of different people are often inconsistent.

In addition, the subjective evaluation method can not quantitatively evaluate the noise source. Therefore, acoustic measurement and signal analysis are often used. ?

Sound pressure method:

Near-field measurement is a simple method, which is usually used to find the main noise sources of machines. The specific method is to scan the surface close to the machine with a sound level meter, and determine the position of the noise source from the indicated value of the sound level meter. According to the acoustic principle, the correctness of the near-field measurement method is conditional. The sound level measured by microphone should be mainly caused by a nearby noise source, while other noise sources have no or little influence on the measured value. But the sound field at a certain point will always be mixed by other sound sources nearby, especially in the workshop. Therefore, the near-field measurement method can not provide accurate measurement values. Therefore, this method is usually used for rough location of machine noise sources.

Selection of operation method The selection of operation method is to try to connect or separate the running parts of the machine step by step according to the measurement requirements, and measure the sound level of some parts and their share in the total sound level of the whole machine, so as to determine the main noise source. This method is very useful for noise source identification of complex machinery, especially multi-stage gear transmission machinery. Of course, this method can only be used when all parts of the machine can be out of operation. Methods of noise source identification and location For example, to estimate the noise generated by the motor of the fan and the fan, you can turn off the fan, only start the motor and measure the noise of the motor. According to the principle of sound level superposition, the sound level and spectrum of fan noise can be estimated from the noise level and spectrum of motor and the total noise level and spectrum of fan. When measuring motor noise, the load of the motor should be kept constant. The greater the difference between fan noise and motor noise, the higher the estimation accuracy of fan noise.

When the running state cannot be changed, the selective coverage method is usually used to identify the noise source. This method is to cover all parts of the machine with sound insulation material (lead plate), measure the noise of the uncovered parts, and determine the noise source. The covering layer (sound insulation cover) shall be designed with special doors to ensure that the noise after covering is 10dB lower than that before covering. When measuring the noise of a certain part, covering other parts is equivalent to measuring each independent noise source separately. By comparing the noise measured in each part, the main noise source can be found. A part of the machine can be covered with a lead plate with the thickness of 1 ~ 1.5 mrn, and the cover is filled with mineral wool or glass fiber. This covering technique can reduce the noise by about 10 ~ 15 DBA, so it is easy to distinguish it from the uncovered vibration surface. But this method is suitable for identifying intermediate frequency and high frequency noise, because the low frequency sound insulation ability of the sound insulation cover is very poor. It can also be distinguished according to noise characteristics. For example, when measuring the mechanical noise and exhaust noise of the engine, the exhaust pipe can be led out of the wall to seal the gap. The mechanical noise of the engine can be measured indoors and the exhaust noise can be measured outside the wall.

Sound intensity method:

In the three-dimensional fluid sound field, the sound intensity vector is equal to the vector sum of the effective sound intensity vector and the sound intensity deviation. The sound intensity deviation represents the acoustic energy flow in the local area of the sound field, and its vector streamline is annular. The deviation of sound intensity in narrow frequency domain is usually a non-zero rotation vector, so the sound intensity vector in narrow frequency band does not necessarily deviate from the sound source in the radial direction. The streamline of sound intensity vector at each frequency point is usually curved, especially in the near field or the area with strong reflected wave, the curvature radius of the streamline is small, and the sound intensity vector at some frequency points even points to the sound source, which shows that the direction of the sound source cannot be inferred from the single-frequency sound intensity vector at several points in the sound field. With the increase of frequency bandwidth, the influence of sound intensity deviation decreases. When the deviation value of sound intensity can be ignored, the sound intensity vector is equal to the effective sound intensity vector. The sound intensity vector streamline represents the actual power streamline in the sound field, that is, from the sound source to the infinite area or ending at the power absorption point. In this case, the direction of the sound source can be judged according to several sound intensity vectors that are not in a plane. Generally, the analysis frequency bandwidth used for sound source localization should not be narrower than 1/3 octave bandwidth; According to experience, it is best to choose a frequency band containing several octaves as the analysis frequency bandwidth. The sound intensity vector of a point is estimated by measuring the sound intensity in three orthogonal directions. For example, in Cartesian coordinate space, if the measured values of sound intensity on three orthogonal axes are Ix, Iy and Iz, the amplitude of sound intensity vector is: the method of noise source identification and location. Usually, it is very time-consuming to locate the sound source with sound intensity technology. Unless the sound intensity meter can measure three orthogonal axial components of the sound intensity vector at the same time, it will make three measurements at each point to determine its sound intensity vector. The accuracy of sound source localization is mainly related to the characteristics of fluid sound field. For resistive sound field, the accuracy of sound source localization is usually high. When the sound intensity vectors of a few points are used to locate the sound source, the positioning accuracy is related to the location selection of the measuring points. The positions of measuring points should be evenly distributed around the sound source. Once the position of the sound source is preliminarily determined, the sound intensity vector at the measuring point far away from the sound source should be abandoned. If the spatial distribution of sound intensity vector in the sound field has been determined, the positions of sound source and power absorption point can be easily determined. Sound intensity technology can also be very effectively used to find the sound leakage position of partition walls or enclosed spaces, and to check the sound insulation quality of enclosed spaces such as sound insulation rooms, anechoic rooms and sound insulation enclosures. Before the sound insulation experiment, the sound intensity technology can be used to check the sealing condition of the test parts. When the sound field is the superposition of radiation fields of several sound sources, the main radiation sound source can be found by sound intensity technology; Arrange sound sources in the order of radiated sound power. For the acoustic radiation of complex machines, the scanning measurement method can be used to measure the acoustic radiation power of various parts (surfaces) of the machine and find out the main acoustic radiation areas or components. We know that in the radiation near field of point sound source or its combination sound source, the reactive component of transient sound intensity is much greater than its active component. But the reverse is not necessarily true, that is, when there is a strong reactance component of transient sound intensity near the surface of an object, it does not mean that the object is the sound source. For example, in a reverberation sound field in a closed room. In addition, the reactance component of near-field transient sound intensity can not reflect the radiation efficiency of sound source. Therefore, the reactance component of transient sound intensity (imaginary part of complex sound intensity) can only be used as an auxiliary means of sound source localization for preliminary analysis.

Array method:?

Microphone array is an array composed of many microphones arranged in a certain way, which has strong directivity and can be used to determine the spatial distribution of sound sources, that is, to find out the position and intensity of sound sources, so it can identify the noise sources when the locomotive is running. Applying digital technology to prestige telescope can realize automatic scanning of acoustic telescope. Therefore, we can analyze high-speed moving sound sources (such as trains, airplanes, etc.) and analyze the frequency spectrum of the received sound signals, so as to obtain the spatial distribution of sound sources in different frequency bands. At present, the most widely used method is to arrange microphones in a straight line. This system is called linear array directivity system. The linear array uses the directivity generated by the interference effect of the received signal at multiple pickup points. However, the sidelobe of this kind of linear array with equal spacing and intensity is relatively large. If the signals of each microphone are corrected according to certain rules, the sidelobe can be suppressed. The commonly used microphone array is modified according to the coefficient of Chebyshev series. This can widen the main lobe, but reduce the sidelobe by 30dB. Microphone array can be completed by analog circuit, but it is generally processed by digital method at present. The output signal of the microphone is sampled and sent to the computer through analog-to-digital conversion. The position of the focus is automatically changed by the computer, and the sound source intensity distribution on the xy line is obtained by scanning. At the same time, the frequency spectrum of each point is calculated by fast Fourier transform. Linear array microphone can only measure sound sources distributed on one line at a time. If you want to analyze the sound source distribution in several directions at the same time, you must use several microphone arrays or squares. Another principle of microphone array telescope is: firstly, the output signals of two microphones in acoustic telescope are cross-correlated, and then the frequency spectrum is obtained by fast Fourier transform with time delay. The frequency spectrum is related to the distance between the two microphones, and the intensity relationship of different frequency bands of sound waves from different directions can be obtained by fast Fourier transform with the distance between the two microphones. ?

Signal analysis time domain analysis?

According to the different time characteristics of each sound source or each part of the sound source, it is more suitable for signals with discrete spectrum. If the machine produces impulse noise, the time history of the noise can be recorded. It is displayed on the bilinear oscilloscope, and the other way is used to display the marking pulse, which is triggered by the moving parts of the machine, so that the noise is related to the mechanical action. Once the noise signal is associated with mechanical vibration, it can be determined that the noise comes from the vibration part. The averaging technique is the development of time domain analysis. Sometimes in the time course of noise and vibration, it is difficult to distinguish scattered and repeated events because of the high background noise. The background noise is segmented according to the working cycle of the machine, and the signals of multiple cycles are averaged. The aperiodic signal grows slowly after averaging for many times, while the periodic signal grows rapidly, so the periodic signal can be detected. Usually, the signals with the duty cycle of 10 ~ 100 are averaged to clearly distinguish the repeated events. The averaging process uses a computer to complete the frequency domain analysis. If the noise of the noise source is in different frequency regions, narrow-band spectrum analysis can be used. The vibration of noise source is measured by accelerometer, and the sound pressure at a certain point is measured by microphone, and their frequency spectrum is obtained for analysis. The main part of the vibration signal spectrum of the noise source and the main part of the sound signal spectrum are located in the same frequency region, or have peaks at some frequencies, so that the noise source can be regarded as the main noise source.