Speed reducer squeal traceability method based on multi-channel sound and vibration information fusion

Abstract

The invention provides a multichannel sound and vibration information fusion speed reducer squeal traceability method, which comprises the following steps of: arranging a plurality of vibration sensors and microphones around a speed reducer, and synchronously acquiring vibration signals and sound signals; converting the multi-channel acoustic vibration signals from a time domain to a frequency domain by adopting short-time Fourier transform and averaging the multi-channel acoustic vibration signals along a time dimension; setting an outburst ratio threshold value to realize howling frequency band screening of the sound spectrum of each channel; summing the screened sound frequency spectrum and vibration frequency spectrum along the channel dimension and carrying out amplitude normalization processing; the processed sound frequency spectrum and the vibration frequency spectrum are multiplied according to corresponding frequency points, and a fusion frequency spectrum of sound and vibration information fusion is obtained; and converting the frequency domain into an order domain to obtain a howling traceability result of the speed reducer. According to the method, after the outburst ratio threshold value is set, professional technicians do not need to analyze data, mutual verification of sound and vibration signals can be achieved, interference of external noise is restrained, fusion of sound and vibration information at the feature level is adopted, and accurate howling traceability can be achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of reducer howling source tracing analysis in a new energy vehicle transmission system, and specifically relates to a reducer howling source tracing method based on multi-channel acoustic and vibration information fusion. Background Art

[0002] The new energy trend is gradually changing the industry landscape of construction machinery and heavy vehicles, among which high-speed and precise electric drive reducers are the core components of these equipment. This type of reducer is characterized by high-speed operation. Under high-speed and high-load working conditions, the lack of noise masking by the engine makes the noise of the electric drive reducer more obvious. In addition, the natural frequency concentration of the transmission system, the complex local modes and the increased resonance points also aggravate the noise problem, which has a serious impact on driving comfort. In particular, the howling sound, as a major form of noise in the electric drive reducer, needs to be traced back to the source analysis in a targeted manner.

[0003] At present, there are two main ways to trace the source of reducer howling: one is to collect sound signals at different positions of the reducer and generate a Campbell diagram to analyze the reducer noise; the other is to trace the source of the howling by blindly separating the multi-channel sound signals. The method of noise analysis through Campbell diagram can only show the distribution of all noise orders and energy. There is no measurement standard for howling, so the source of howling cannot be clearly determined. The analysis method is complicated and depends on the experience of professionals. The method of tracing the howling source through blind source separation will cause a jump in the tracing result due to the reduction in the number of channels, making the tracing result non-unique. Summary of the invention

[0004] The purpose of the present invention is to solve the problems in the prior art that the reducer howling tracing method needs to conduct targeted analysis on the results and the tracing results are uncertain, and to provide a reducer howling tracing method with multi-channel acoustic and vibration information fusion.

[0005] To achieve the above purpose, the technical solution provided by the present invention is:

[0006] A method for tracing the source of reducer howling by fusion of multi-channel acoustic and vibration information, comprising the following steps:

[0007] Step 1: Arrange multiple vibration sensors and multiple microphones around the reducer to synchronously collect vibration signals and sound signals at various positions and directions of the reducer;

[0008] Step 2, respectively converting the collected multi-channel sound signal and vibration signal from the time domain to the frequency domain using short-time Fourier transform, and averaging the obtained spectrum results along the time dimension;

[0009] Step 3, setting a protrusion ratio threshold, converting the spectrum result of the multi-channel sound signal obtained in step 2 into a sound pressure level, and calculating the protrusion ratio at each frequency point, retaining the amplitude of the intermediate bandwidth frequency range corresponding to the protrusion ratio greater than the protrusion ratio threshold;

[0010] Step 4, summing the multi-channel acoustic signal spectrum obtained in step 3 and the multi-channel vibration signal spectrum obtained in step 2 along the channel dimension, and performing amplitude normalization processing on each of them;

[0011] Step 5, multiplying the sound spectrum and the vibration spectrum obtained in step 4 according to the corresponding frequency points to obtain a fused spectrum after fusion of multi-channel sound and vibration information;

[0012] Step 6: Divide the frequency coordinate of the fused spectrum by the speed of the reducer input shaft to obtain the howling source tracing result in the order domain.

[0013] Furthermore, in step 1, six vibration sensors are arranged on the reducer surface near the bearing position, and four microphones are dispersedly arranged at a distance of 1000 mm from the reducer surface.

[0014] Furthermore, in step 2, the window function used by the short-time Fourier transform is a Hanning window, and the window length is 65536.

[0015] Further, in step 3, the protrusion ratio threshold is set to 15 dB, and the frequency range for calculating the protrusion ratio is 90 Hz to 7284 Hz.

[0016] Furthermore, in step 4, the summation result of the vibration signal and the summation result of the acoustic signal are normalized to [0, 1].

[0017] The advantages of the present invention are:

[0018] 1. The multi-channel sound and vibration information fusion reducer howling source tracing method of the present invention adopts an end-to-end reducer howling source tracing method. After setting the protrusion ratio threshold, there is no need for professional technicians to analyze the data. In addition, since multiple vibration sensors are used to synchronously collect the reducer vibration signal, the frequency that exists in the sound signal but not in the vibration signal can be considered as an external interference signal, which realizes the mutual verification of the sound and vibration signals and suppresses the interference of external noise. It also adopts the fusion of sound and vibration information at the feature level, and there is no information loss. The vibration signal plays a role in further verifying the sound signal, which can realize accurate howling source tracing, thereby solving the problems of complex analysis method, reliance on professional experience and uncertain tracing results of existing howling source tracing methods.

[0019] 2. The spectrum synthesis of the acoustic vibration signals along the channel dimension is adopted, which effectively takes into account the different acoustic vibration responses in different directions. The proportion of the energy of the common frequency components of each channel in the overall frequency energy is enhanced, and the proportion of the energy of the non-common frequency components in the overall frequency energy is weakened, thereby achieving the enhancement of the common components in the acoustic vibration signals and improving the accuracy of the tracing results. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The above and / or other features and advantages of the present invention will become more easily understood through the following description with reference to the accompanying drawings, which are not drawn to scale and some features are exaggerated or reduced to show details of specific components. In the accompanying drawings:

[0021] Figure 1 It is a flow chart of the method for tracing the source of reducer howling by fusion of multi-channel acoustic and vibration information of the present invention;

[0022] Figure 2 is the calculation result of the protrusion ratio in the example of the present invention;

[0023] Figure 3 This is a diagram of the reducer howling tracing result in an example of the present invention. DETAILED DESCRIPTION

[0024] The present invention will be described in detail below with reference to the accompanying drawings by means of exemplary embodiments of the present invention. It should be noted that the following detailed description of the present invention is only for the purpose of illustration, and is not intended to limit the present invention.

[0025] The present invention provides a method for tracing the howling of a reducer by fusing multi-channel acoustic and vibration information. In the tracing of the howling of the reducer, multi-channel vibration signals and multi-channel acoustic signals are used. The vibration signals and acoustic signals are respectively converted from the time domain to the frequency domain by short-time Fourier transform, and are averaged according to the time dimension to achieve noise reduction. A prominent ratio threshold is used to achieve the howling frequency band screening of the acoustic spectrum of each channel. The screened acoustic spectrum and vibration spectrum are synthesized in the channel dimension according to their respective categories, and the synthesized acoustic and vibration spectra are normalized respectively. The processed acoustic and vibration spectra are multiplied according to corresponding frequency points to achieve the fusion of acoustic and vibration information and obtain the howling frequency domain result. The frequency domain is converted to the order domain to achieve the tracing of the howling of the reducer.

[0026] Reference Figure 1 As an exemplary embodiment of the present invention, a method for tracing the source of reducer howling by fusion of multi-channel acoustic and vibration information includes the following steps:

[0027] Step S1, arranging a plurality of vibration sensors and a plurality of microphones around the reducer to synchronously collect vibration signals and sound signals at various positions and directions of the reducer;

[0028] Step S2, respectively converting the collected multi-channel sound signal and vibration signal from the time domain to the frequency domain using short-time Fourier transform, and averaging the obtained spectrum results along the time dimension;

[0029] Step S3, setting a protrusion ratio threshold, converting the spectrum result of the multi-channel sound signal obtained in step S2 into a sound pressure level, and calculating the protrusion ratio at each frequency point, retaining the amplitude of the intermediate bandwidth frequency range corresponding to the protrusion ratio greater than the protrusion ratio threshold;

[0030] Step S4, summing the multi-channel acoustic signal spectrum obtained in step S3 and the multi-channel vibration signal spectrum obtained in step S2 along the channel dimension, and performing amplitude normalization processing on each of them;

[0031] Step S5, multiplying the sound spectrum and the vibration spectrum obtained in step S4 according to the corresponding frequency points to obtain a fused spectrum after fusion of multi-channel sound and vibration information;

[0032] Step S6, dividing the frequency coordinate of the fused spectrum by the speed of the reducer input shaft to obtain the howling source tracing result in the order domain.

[0033] In step S1, multiple vibration sensors and microphones can be arranged on the reducer surface near the bearing position and at a spatial position 1 meter away from the reducer, respectively. The number of vibration sensors and microphones is not particularly limited. In particular, six vibration sensors can be arranged on the reducer surface near the bearing position, and four microphones can be dispersedly arranged 1000mm away from the reducer surface, in the directions of left, top, right, and upper left, respectively, to collect synchronized multi-channel acoustic and vibration signals of the reducer under steady-state working speed and load. Since multiple vibration sensors are used to synchronously collect the vibration signal of the reducer, the frequencies that exist in the acoustic signal but not in the vibration signal can be considered as external interference signals, which effectively solves the problem that the tracing result of howling is easily interfered by external noise, thereby realizing mutual verification of acoustic and vibration signals and suppressing interference from external noise.

[0034] In step S2, the collected multi-channel acoustic vibration signals of the reducer are respectively subjected to short-time Fourier transform, the window function can select the Hanning window, the window length can be 65536, and the spectrum results obtained from each channel are averaged along the time dimension to suppress interference noise.

[0035] The protrusion ratio threshold value is set in step S3 to implement the howling frequency band screening of the sound spectrum of each channel, convert the spectrum results corresponding to the multi-channel sound signal into the sound pressure level, calculate the protrusion ratio at each frequency point of the sound signal of each channel according to the ECMA-418-1 standard, and retain the intermediate frequency band corresponding to the frequency point greater than the protrusion ratio threshold value. The protrusion ratio threshold value can be set to 15dB, and the frequency range for calculating the protrusion ratio is 90Hz to 7284Hz. The corresponding protrusion ratio can be calculated according to the frequency interval, and the intermediate frequency band corresponding to the frequency with a protrusion ratio greater than the 15dB threshold value is retained.

[0036] Step S4 can synthesize the processed vibration signal spectrum results and the screened sound signal spectrum results along the channel dimension, respectively, to obtain the results of multi-channel vibration information fusion and multi-channel sound information fusion, and perform amplitude normalization on the two. The summed results of the vibration signal and the sound signal can be normalized to [0,1], but this does not limit the present invention, and other normalization methods can also be used. Since the spectrum synthesis of the acoustic vibration signals along the channel dimension is adopted, the problem of different acoustic vibration responses in different directions is effectively considered, the proportion of the energy of the common frequency components of each channel in the overall frequency energy will be strengthened, and the proportion of the energy of the non-common frequency components in the overall frequency energy will be weakened, thereby achieving the enhancement of the common components in the acoustic vibration signal and improving the accuracy of the tracing results.

[0037] In step S5, since the vibration signal and the acoustic signal have the same sampling rate and the number of points for Fourier transform is the same, the spectrum results of the acoustic vibration signal have the same frequency resolution, and the spectrum results of the acoustic vibration signal of the corresponding frequency can be directly multiplied to obtain the result of multi-channel acoustic vibration information fusion. Since the acoustic vibration information is fused at the feature level (referring to the frequency in the processing process is the characteristic of the signal), the vibration signal plays a further verification role of the acoustic signal in the absence of information loss, effectively solving the problem of uncertain tracing results, and thus achieving accurate tracing of the howling source.

[0038] Step S6 is used to convert the spectrum result obtained by fusion of multi-channel acoustic and vibration information from the frequency domain to the order domain, and realize the tracing of the reducer howling by fusion of multi-channel acoustic and vibration information when the structure of the reducer transmission system is known. In this embodiment, the frequency coordinate of the spectrum after fusion of multi-channel acoustic and vibration information is divided by the speed of the reducer input shaft to obtain the tracing result converted from the frequency domain to the order domain.

[0039] As described above, the present invention adopts an end-to-end reducer howling tracing method. After setting the protrusion ratio threshold, there is no need for professional technicians to analyze the data. In addition, since multiple vibration sensors are used to synchronously collect the reducer vibration signal, the frequency that exists in the sound signal but not in the vibration signal can be considered as an external interference signal, thereby realizing mutual verification of the sound and vibration signals and suppressing the interference of external noise. It also adopts the fusion of sound and vibration information at the feature level, and there is no information missing. The vibration signal plays a further verification role of the sound signal, which can achieve accurate howling tracing, thereby solving the problems of complex analysis methods, reliance on professional experience and uncertain tracing results of existing howling tracing methods.

[0040] Next, the reducer howling source tracing method based on multi-channel acoustic and vibration information fusion provided by the present invention is further explained with reference to examples.

[0041] In this example, the object of collection is the FS4E200-2 electric drive reducer, and the multi-channel acoustic and vibration signals of the reducer are collected synchronously at the third gear, the load is 240Nm, and the input shaft speed is constant at 4200rpm. Six vibration sensors are arranged on the reducer surface near the bearing position, and four microphones are arranged on the left, top, right, and upper left at a distance of 1000mm from the reducer surface.

[0042] After collecting multi-channel sound signals and vibration signals, short-time Fourier transform is used to convert the collected sound signals and vibration signals from time domain to frequency domain, and the obtained spectrum results are averaged along the time dimension, where the window function is Hanning window and the window length is 65536.

[0043] Then, the prominence ratio threshold is set to 15dB, and the prominence ratios of the four channels of the sound signal from 90Hz to 7284Hz are calculated. The calculation results of the prominence ratios of the four-channel sound signals are as follows: Figure 2 The middle frequency band corresponding to the frequencies with prominence ratio greater than the 15dB threshold is retained.

[0044] Then, the vibration signal spectra of the six channels and the four acoustic signal spectra after threshold screening are summed up according to the corresponding frequencies, and the results of the vibration signal fusion and the acoustic signal fusion are normalized to [0, 1].

[0045] Finally, the spectral frequency coordinates of the fused multi-channel acoustic and vibration information are divided by the speed of the reducer input shaft to obtain the traceability result converted from the frequency domain to the order domain, such as Figure 3 As shown. Figure 3It can be seen that at 240Nm-4200rpm in the 3rd gear, the traceability results obtained are sorted by energy as 9.696 and 35.866, respectively. The order with the largest energy, that is, the largest contribution, is 9.696, and the reducer has 36, 23.774, 13.868 and 9.732 orders of gear pair meshing frequency corresponding to the order in the 3rd gear. Therefore, it can be seen that the main source of the howling is the gear pair corresponding to the 9.732 order, and the difference between the traceability order and the inherent order of the reducer is mainly due to the fact that the speed is not the ideal 4200rpm. According to the traceability results, the corresponding gear pair in the reducer transmission system can be optimized to reduce the howling noise of the reducer. Therefore, the applicability and effectiveness of the reducer howling traceability method with multi-channel acoustic and vibration information fusion provided by the present invention are verified through experiments.

[0046] Finally, it should be noted that the features mentioned and / or shown in the above description of the exemplary embodiments of the present invention may be combined in the same or similar manner into one or more other embodiments, combined with the features in other embodiments or substituted for the corresponding features in other implementations. The technical solutions obtained by these combinations or substitutions shall also be deemed to be included in the protection scope of the present invention.

Claims

1. A method for tracing the source of reducer howling by fusion of multi-channel acoustic and vibration information, characterized in that: The following steps are involved: Step 1: Arrange multiple vibration sensors and multiple microphones around the reducer to synchronously collect vibration signals and sound signals at various positions and directions of the reducer; Step 2, respectively converting the collected multi-channel sound signal and vibration signal from the time domain to the frequency domain using short-time Fourier transform, and averaging the obtained spectrum results along the time dimension; Step 3, setting a protrusion ratio threshold, converting the spectrum result of the multi-channel sound signal obtained in step 2 into a sound pressure level, and calculating the protrusion ratio at each frequency point, retaining the amplitude of the intermediate bandwidth frequency range corresponding to the protrusion ratio greater than the protrusion ratio threshold; Step 4, summing the multi-channel acoustic signal spectrum obtained in step 3 and the multi-channel vibration signal spectrum obtained in step 2 along the channel dimension, and performing amplitude normalization processing on each of them; Step 5, multiplying the sound spectrum and the vibration spectrum obtained in step 4 according to the corresponding frequency points to obtain a fused spectrum after fusion of multi-channel sound and vibration information; Step 6: Divide the frequency coordinate of the fused spectrum by the speed of the reducer input shaft to obtain the howling source tracing result in the order domain.

2. The method for tracing the source of reducer howling by multi-channel acoustic and vibration information fusion according to claim 1 is characterized by: In step 1, six vibration sensors are arranged on the reducer surface near the bearing position, and four microphones are dispersedly arranged at a distance of 1000 mm from the reducer surface.

3. The method for tracing the source of reducer howling by multi-channel acoustic and vibration information fusion according to claim 1 or 2, characterized in that: In step 2, the window function used for short-time Fourier transform is the Hanning window with a window length of 65536.

4. The method for tracing the source of reducer howling by multi-channel acoustic and vibration information fusion according to claim 1 or 2, characterized in that: In step 3, the protrusion ratio threshold is set to 15 dB, and the frequency range for calculating the protrusion ratio is 90 Hz to 7284 Hz.

5. The method for tracing the source of reducer howling by multi-channel acoustic and vibration information fusion according to claim 1 or 2, characterized in that: In step 4, the summation result of the vibration signal and the summation result of the acoustic signal are normalized to [0, 1].

Images