Processing method and device for eliminating abnormal aerodynamic noise of model rotor caused by external airflow disturbance

Abstract

The invention discloses a processing method and device for eliminating abnormal aerodynamic noise of a model rotor caused by external airflow disturbance, and the method comprises the steps: 1, segmenting the collected aerodynamic noise data of the rotor according to circles according to the characteristic that the aerodynamic noise of the rotor in a stable state has dominant frequency stability; 2, performing amplitude and frequency analysis on the obtained data of each circle, and extracting frequency values corresponding to the first three maximum amplitudes in the amplitude spectrum of the data of each circle; 3, performing statistics on the frequency values extracted from all circles to obtain the main frequency of the most public number; and step 4, eliminating data of circles corresponding to main frequencies of non-most people, carrying out whole-period averaging on the data of the remaining circles, and judging the confidence level of the noise data. The technical scheme provided by the embodiment of the invention solves the problems that the existing rotor aerodynamic noise test mode has high requirements on the test environment and the test site, so that the test environment and the test site limit the existing test bed to carry out related noise tests; therefore, a rotor aerodynamic noise test is difficult to carry out in an environment with external gust interference.

Description

Technical Field

[0001] The present invention relates to the field of, but is not limited to, rotor noise testing technology for rotorcraft, and in particular to a method and device for eliminating abnormal aerodynamic noise of a model rotor caused by external airflow disturbance. Background Art

[0002] Rotor aerodynamic noise is the main source of external noise for rotorcraft. Studying the acoustic field distribution characteristics of rotor aerodynamic noise has guiding significance for whether rotorcraft can pass noise airworthiness certification. Reducing rotor aerodynamic noise is the main way to reduce the external noise of rotorcraft. Since rotor aerodynamic noise has the characteristics of low frequency, long propagation distance and slow attenuation, effective measurement of it can provide first-hand experimental data for rotor noise reduction research.

[0003] The rotor aerodynamic noise test research of rotorcraft is currently mainly carried out in an open space without any obstructions in the outdoor field, or in an anechoic chamber and anechoic wind tunnel with a closed and anechoic environment. In other words, the rotor aerodynamic noise test has high requirements for the test environment and test site, but the requirements of the environment and site also limit the relevant noise tests on the existing test benches. Summary of the invention

[0004] Purpose of the present invention: In order to solve the above-mentioned problem, the embodiments of the present invention provide a method and device for eliminating abnormal aerodynamic noise of a model rotor caused by external airflow disturbance, so as to solve the problem that the existing rotor aerodynamic noise test method has high requirements for the test environment and the test site, so that the test environment and the site limit the existing test bench to carry out related noise tests, resulting in the problem that the rotor aerodynamic noise test is difficult to carry out in an environment with external gust interference.

[0005] Technical solution of the present invention: In a third aspect, an embodiment of the present invention provides a method for eliminating abnormal aerodynamic noise of a model rotor caused by external airflow disturbance, comprising:

[0006] Step 1, according to the characteristic that the rotor aerodynamic noise in a stable state has a main frequency stability, the collected rotor aerodynamic noise data is divided into circles;

[0007] Step 2, performing amplitude and frequency analysis on the obtained data of each circle, and extracting the frequency values ​​corresponding to the first three maximum amplitudes in the amplitude spectrum of each circle of data;

[0008] Step 3, count the frequency values ​​extracted from all circles to obtain the main frequency with the largest number of frequencies;

[0009] Step 4: Eliminate the data of the circles corresponding to the main frequency that is not the most common, average the data of the remaining circles over the entire period, and determine the confidence level of the noise data.

[0010] Optionally, in the above-mentioned method for removing abnormal aerodynamic noise of a model rotor caused by external airflow disturbance, step 1 comprises:

[0011] Step 11, synchronously collecting noise data and rotor speed pulse data of each measuring point at a sampling frequency St;

[0012] Step 12, according to the sampling frequency St and the rotor speed Nr, calculate the number of data points per rotor rotation Np = St / Nr;

[0013] Step 13: Select the noise data of the measurement point farthest from the center line of the model rotor shaft from each noise measurement point, and divide M into the measurement points M and M in sequence with any pulse rising edge of the rotor speed Nr as the starting point. n The noise data sequence has Np data points in each circle, that is, M n 1-dimensional arrays, each of which has a length of Np.

[0014] Optionally, in the above-mentioned method for removing abnormal aerodynamic noise of the model rotor caused by external airflow disturbance, in step 11,

[0015] The collected rotor speed pulse data is a signal that changes from a low level to a high level every time the rotor rotates one circle, the high level lasts for a period of time, and then returns to a low level. The duty cycle of the high level duration period of each circle ranges from 20% to 30%, St≥1024×Nr; where Nr is the rotor speed;

[0016] Among them, the number of rising edge data points is 4 to 10, and the time length T of data collection satisfies:

[0017] T = M0 × Nr;

[0018] M0 is the number of turns, and the value range of M0 is 200≥M0≥130.

[0019] Optionally, in the above-mentioned method for removing abnormal aerodynamic noise of the model rotor caused by external airflow disturbance, the manner of performing amplitude and frequency analysis on each circle of data in step 2 includes:

[0020] Step 21: M obtained by segmentation in step 1 n The 1-dimensional arrays are analyzed for auto-power spectrum in turn, so as to obtain the amplitude 1-dimensional array and frequency 1-dimensional array of the auto-power spectrum of each circle of data;

[0021] Among them, the one-dimensional array of the amplitude of the power spectrum of the first cycle of data is marked as [A 1-1 , A 1-2 …、A 1-Np / 2 ] and the one-dimensional array of frequency values ​​is identified as [F 1-1 、F1-2 …、F 1-Np / 2 ], No. M n The one-dimensional array of the amplitude of the power spectrum of the loop data is marked as [A Mn-1 , A Mn-2 …、A Mn-Np / 2 ] and the one-dimensional array of frequency values ​​is identified as [F Mn-1 、F Mn-2 …、F Mn-Np / 2 ].

[0022] Optionally, in the above-mentioned method for removing abnormal aerodynamic noise of the model rotor caused by external airflow disturbance, the method of extracting the frequency values ​​corresponding to the first three maximum amplitudes in the amplitude spectrum of each circle of data in step 2 includes:

[0023] Step 22, according to the one-dimensional array of amplitudes of the power spectrum of each circle of data obtained in step 21, extract three frequency values ​​corresponding to the first three maximum amplitudes in the one-dimensional array of amplitudes of the circle;

[0024] Among them, the frequency values ​​corresponding to the first three maximum amplitudes in the one-dimensional array of the first circle amplitude are marked as [F max-1-1 ,F max-1-2 ,F max-1-3 ], No. M n The frequency values ​​corresponding to the first three largest amplitudes in the one-dimensional array of circle amplitudes are marked as [F max-Mn-1 ,F max-Mn-2 ,F max-Mn-3 ].

[0025] Optionally, in the above-mentioned method for removing abnormal aerodynamic noise of the model rotor caused by external airflow disturbance, step 3 comprises:

[0026] According to the sorting of amplitude size, for the frequency data of all circles, first extract the maximum mode and number of rows of the frequency corresponding to the first amplitude value, and then extract the maximum mode and number of rows of the frequency corresponding to the second amplitude value based on the maximum mode row corresponding to the first amplitude value, and then extract the maximum mode and number of rows of the frequency corresponding to the third amplitude value based on the maximum mode row corresponding to the second amplitude value.

[0027] Optionally, in the above-mentioned method for removing abnormal aerodynamic noise of the model rotor caused by external airflow disturbance, step 3 comprises:

[0028] Step 31, according to the M obtained in step 22 n The frequency values ​​corresponding to the first three maximum amplitudes in each circle generate 4 columns M of the following form n Two-dimensional array M of rows all Identified as:

[0029] [F max-1-1 ,Fmax-1-2 ,F max-1-3 ,1;

[0030] F max-2-1 ,F max-2-2 ,F max-2-3 ,2;

[0031] ……;

[0032] F max-Mn-1 ,F max-Mn-2 ,F max-Mn-3 ,M n ];

[0033] Step 32, from the two-dimensional array M all Select the largest common factor F in the first column max-1 , and the number of rows with the largest common number is N max-1 ;

[0034] Step 33, from the two-dimensional array M all Extract the first number in each row equal to F max-1 All rows form a new 4-column N max-1 Two-dimensional array M of rows n-max-1 ;

[0035] Step 34, from the two-dimensional array M n-max-1 Select the largest common factor F in the second column max-2 , and the number of rows with the largest common number is N max-2 ;

[0036] Step 35, from the two-dimensional array M n-max-1 Extract the second number in each row equal to F max-2 All rows form a new 4-column N max-2 Two-dimensional array M of rows n-max-2 ;

[0037] Step 36, from the two-dimensional array M n-max-2 Select the largest common factor F in the third column max-3 , and the number of rows with the largest common number is N max-3 ;

[0038] Step 37, from the two-dimensional array M n-max-2 Extract the third number in each row equal to F max-3 All rows form a new 4-column N max-3 Two-dimensional array M of rows n-max-3 .

[0039] Optionally, in the above-mentioned method for removing abnormal aerodynamic noise of the model rotor caused by external airflow disturbance, step 4 includes the following situations:

[0040] If Nmax-1 <50%M n , then the test data is confirmed to be invalid;

[0041] If 50% M n ≤N max-1 ≤70%M n , then press M n-max-1 The fourth number in each row is used to extract the corresponding circle and average the whole cycle. The averaged data is used for trend analysis and quantitative analysis with low confidence level.

[0042] If N max-1 >70%M n , and N max-2 ≤60%M n , then press M n-max-3 The fourth number in each row is used to extract the corresponding circle for full-cycle averaging, and the test data of this round is considered valid. The averaged data is used for trend analysis and quantitative analysis with medium confidence.

[0043] If N max-1 >70%M n , and N max-3 ≤50%M n , then press M n-max-3 The fourth number in each row is used to extract the corresponding circle for full-cycle averaging, and the test data of this round is considered valid. The averaged data is used for trend analysis and quantitative analysis with medium confidence.

[0044] If N max-1 >70%M n , and N max-2 >60%M n And N max-3 >50%M n , then press M n-max-3 The fourth number in each row is used to extract the corresponding circle for full-cycle averaging, and the test data of this round is considered to be completely valid. The averaged data is used for trend analysis and quantitative analysis with high confidence.

[0045] In a second aspect, an embodiment of the present invention provides a device for processing abnormal aerodynamic noise of a model rotor caused by external airflow disturbance, comprising: a data acquisition and segmentation module, an extraction module, a statistics module, and a data processing and identification module;

[0046] The data collection and segmentation module is used to segment the collected rotor aerodynamic noise data by circles according to the characteristic that the rotor aerodynamic noise has a main frequency stability in a stable state;

[0047] An extraction module is used to perform amplitude and frequency analysis on the obtained data of each circle, and extract the frequency values ​​corresponding to the first three maximum amplitudes in the amplitude spectrum of each circle of data;

[0048] The statistical module counts the frequency values ​​extracted from all circles to obtain the main frequency with the largest number of frequencies;

[0049] The data processing and identification module removes the data of the circles corresponding to the main frequency that is not the most common, averages the data of the remaining circles over a whole period, and identifies the confidence level of the noise data.

[0050] In a third aspect, an embodiment of the present invention provides a computer-readable storage medium, including: a memory and a processor;

[0051] The memory is configured to store executable instructions;

[0052] The processor is specifically configured to implement a method for processing abnormal aerodynamic noise of a model rotor caused by eliminating external airflow disturbances as described in any one of the above items when executing the executable instructions stored in the memory.

[0053] Beneficial effects of the present invention: An embodiment of the present invention provides a method for processing abnormal aerodynamic noise of a model rotor caused by eliminating external airflow disturbances. Based on the requirement of eliminating disturbance signals in a rotor aerodynamic noise test, a method and device for processing abnormal aerodynamic noise of a model rotor according to an embodiment of the present invention are proposed. According to the characteristic that the rotor aerodynamic noise has main frequency stability in a stable state, the collected rotor aerodynamic noise data is divided into whole cycles, and then the amplitude and frequency analysis is performed on each circle of data obtained, and the frequency values ​​corresponding to the top three maximum amplitudes in the amplitude spectrum of each circle of data are extracted. The frequency values ​​extracted from all circles are counted to obtain the main frequency with the largest number of frequencies, and the data of the circles corresponding to the main frequencies not with the largest number of frequencies are eliminated. Then, the data of the remaining circles are averaged over the whole cycles, and the confidence level of the noise data is identified, thereby achieving effective elimination of the abnormal aerodynamic noise of the rotor caused by sudden gusts, thereby ensuring that effective rotor rotation noise test data can be extracted from the abnormal aerodynamic noise of the rotor caused by external airflow disturbances in an open environment. By adopting the technical solution provided by the embodiment of the present invention, the process of continuous change of the rotor shaft tilt angle can be accurately simulated in a set of equations, and the dynamic response of the tilt-rotor helicopter in the transition state can be solved therefrom. Specific implementation plan

[0054] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only a part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0055] As described in the above background technology, the rotor aerodynamic noise test method has high requirements for the test environment and test site, which limits the existing test benches from carrying out related noise tests, resulting in the rotor aerodynamic noise test being difficult to carry out in an environment with external gust interference.

[0056] In this field, patent CN108051659A discloses a method for separating and extracting rotor noise. The patent method first obtains the rotor noise fundamental frequency, engine noise fundamental frequency and tail rotor noise fundamental frequency of the helicopter, then converts the noise signal into a frequency domain signal by Fourier transform, selects a cutoff frequency for high-pass filtering, removes environmental noise, engine noise and tail rotor noise from the frequency domain signal, and then performs an inverse Fourier transform to obtain the sound pressure time domain signal of the rotor noise. The present invention is aimed at removing the rotor disturbance noise caused by sudden gusts of wind from the outside, and it is performed in the time domain, which is significantly different from the method described in patent CN108051659A. Patent CN202011309542.9 discloses a whole-cycle averaging method for rotor vortex interference noise time domain data based on peak search. This method targets vortex interference noise. The collected test bench rotor vortex interference noise data is intercepted for a whole cycle, and then the data lengths of the intercepted multiple cycles are statistically analyzed. The whole-cycle data with the same data length and the largest proportion are averaged to obtain the required rotor vortex interference noise time domain average data. This method does not take into account the disturbance effect of external gusts on normal rotor noise.

[0057] Theoretically, the rotor aerodynamic noise test in the hovering state should be dominated by thickness noise and load noise (and usually the noise of 1 times the passing frequency component should be the largest), that is, the single-circle time history curve after the whole cycle average has obvious single-wave characteristics, and the waveform shape of the measuring points with the same angle in different azimuths should be basically the same except for one azimuth angle (the difference between the two azimuth angles). Most of the waveforms in the data processing results conform to the above rules. However, the rotor hovering performance test is usually carried out in the hovering test room. During the test, in order to ensure the smoothness of the rotor airflow, all the rolling doors in the hovering test room are opened. Therefore, it is impossible to guarantee the additional disturbance caused by external gusts to the normal rotor, and this disturbance will be reflected in the collected rotor aerodynamic noise data. This disturbance is a kind of interference. The disturbance signal must be eliminated to obtain real and effective rotor aerodynamic noise data.

[0058] To solve the above problems, an embodiment of the present invention provides a method and device for processing abnormal aerodynamic noise of a model rotor caused by external airflow disturbances, which can accurately simulate the process of continuous change of the rotor shaft tilt angle in a set of equations, and can solve the dynamic response of the tilt-rotor helicopter in the transition state.

[0059] An embodiment of the present invention provides a method for processing abnormal aerodynamic noise of a model rotor caused by external airflow disturbance. The method for processing abnormal aerodynamic noise of a model rotor comprises the following steps:

[0060] Step 1, according to the characteristic that the rotor aerodynamic noise in a stable state has a main frequency stability, the collected rotor aerodynamic noise data is divided into circles;

[0061] Step 2, performing amplitude and frequency analysis on the obtained data of each circle, and extracting the frequency values ​​corresponding to the first three maximum amplitudes in the amplitude spectrum of each circle of data;

[0062] Step 3, count the frequency values ​​extracted from all circles to obtain the main frequency with the largest number of frequencies;

[0063] Step 4: Eliminate the data of the circles corresponding to the main frequency that is not the most common, average the data of the remaining circles over the entire period, and determine the confidence level of the noise data.

[0064] The method for processing the abnormal aerodynamic noise of a model rotor caused by external airflow disturbances provided in an embodiment of the present invention can effectively eliminate the abnormal aerodynamic noise of the rotor caused by sudden gusts of wind, thereby ensuring that effective rotor rotation noise test data can be extracted from the abnormal aerodynamic noise of the rotor caused by external airflow disturbances in an open environment.

[0065] In one implementation of the embodiment of the present invention, the above step 1 may include:

[0066] Step 11, synchronously collecting noise data and rotor speed pulse data of each measuring point at a sampling frequency St;

[0067] Step 12, according to the sampling frequency St and the rotor speed Nr, calculate the number of data points per rotor rotation Np = St / Nr;

[0068] Step 13: Select the noise data of the measurement point farthest from the center line of the model rotor shaft from each noise measurement point, and divide M into the measurement points M and M in sequence with any pulse rising edge of the rotor speed Nr as the starting point. n The noise data sequence has Np data points in each circle, that is, M n 1-dimensional arrays, each of which has a length of Np.

[0069] In this step, for example, M n ≥100, in the embodiment of the present invention, M n Take 100 as an example to illustrate, that is, 100 one-dimensional arrays are obtained by segmentation, and the length of the one-dimensional array is Np.

[0070] In one implementation of the embodiment of the present invention, the implementation method of the above step 11 is:

[0071] The collected rotor speed pulse data is a signal that changes from a low level to a high level every time the rotor rotates one circle, the high level lasts for a period of time and then returns to a low level. For example, it is similar to a TTL signal. The duty cycle of the high level duration period of each circle ranges from 20% to 30%, and St≥1024×Nr (Nr is the rotor speed, unit: revolutions / second).

[0072] In this implementation, for example, the number of rising edge data points is 4 to 10, and the time length T for collecting data satisfies:

[0073] T = M0 × Nr;

[0074] M0 is the number of turns, and the value range of M0 is 200≥M0≥130.

[0075] In an implementation of the embodiment of the present invention, the method of performing amplitude and frequency analysis on each circle of data in the above step 2 may include:

[0076] Step 21: M obtained by segmentation in step 1 n The one-dimensional arrays (for example, 100 noise data) are subjected to auto-power spectrum analysis in turn (the sampling rate in the auto-power spectrum analysis is set to St), thereby obtaining the one-dimensional array of amplitude and frequency values ​​of the auto-power spectrum of each data circle;

[0077] Among them, the one-dimensional array of the amplitude of the power spectrum of the first cycle of data is marked as [A 1-1 , A 1-2 …、A 1-Np / 2 ] and the one-dimensional array of frequency values ​​is identified as [F 1-1 、F 1-2 …、F 1-Np / 2 ], No. M n The one-dimensional array of the amplitude of the power spectrum of the loop data is marked as [A Mn-1 , A Mn-2 …、A Mn-Np / 2 ] and the one-dimensional array of frequency values ​​is identified as [F Mn-1 、F Mn-2 …、F Mn-Np / 2 ].

[0078] In this implementation, the one-dimensional array of amplitudes of the 100th cycle is labeled [A 100-1 , A 100-2 …、A 100-Np / 2 ] and the one-dimensional array of frequency values ​​is identified as [F 100-1 、F 100-2 …、F 100-Np / 2 ].

[0079] In an implementation of the embodiment of the present invention, the method of extracting the frequency values ​​corresponding to the first three maximum amplitudes in the amplitude spectrum of each circle of data in the above step 2 may include:

[0080] Step 22: generate a one-dimensional array of the amplitude of the power spectrum of each circle of data obtained in step 21 (such as [M 1-1 、M 1-2 …、M 1-(Np / 2) ]), extract the three frequency values ​​corresponding to the first three maximum amplitudes in the one-dimensional array of the circle amplitude;

[0081] Among them, the frequency values ​​corresponding to the first three maximum amplitudes in the one-dimensional array of the first circle amplitude are marked as [F max-1-1 ,F max-1-2 ,F max-1-3 ], No. M n The frequency values ​​corresponding to the first three largest amplitudes in the one-dimensional array of circle amplitudes are marked as [F max-Mn-1 ,F max-Mn-2 ,F max-Mn-3 ].

[0082] In this implementation, for example, the frequency values ​​corresponding to the first three maximum amplitudes in the one-dimensional array of the first cycle amplitudes are identified as [F max-1-1 ,F max-1-2 ,F max-1-3 ], the frequency values ​​corresponding to the first three maximum amplitudes in the one-dimensional array of the 100th amplitude are marked as [F max-100-1 ,F max-100-2 ,F max-100-3 ].

[0083] In an implementation of the embodiment of the present invention, the implementation of the above step 3 may include:

[0084] According to the sorting of amplitude size, for the frequency data of all circles, first extract the maximum mode and number of rows of the frequency corresponding to the first amplitude value, and then extract the maximum mode and number of rows of the frequency corresponding to the second amplitude value based on the maximum mode row corresponding to the first amplitude value, and then extract the maximum mode and number of rows of the frequency corresponding to the third amplitude value based on the maximum mode row corresponding to the second amplitude value.

[0085] In an implementation of the embodiment of the present invention, the specific implementation process of the above step 3 may include:

[0086] Step 31, according to the M obtained in step 22 n The frequency values ​​corresponding to the first three maximum amplitudes in each circle generate 4 columns M of the following form n Two-dimensional array M of rows all Identified as:

[0087] [F max-1-1,F max-1-2 ,F max-1-3 ,1;

[0088] F max-2-1 ,F max-2-2 ,F max-2-3 ,2;

[0089] ……;

[0090] F max-Mn-1 ,F max-Mn-2 ,F max-Mn-3 ,M n ];

[0091] Step 32, from the two-dimensional array M all Select the largest common factor F in the first column max-1 , and the number of rows with the largest common number is N max-1 ;

[0092] Step 33, from the two-dimensional array M all Extract the first number in each row equal to F max-1 All rows form a new 4-column N max-1 Two-dimensional array M of rows n-max-1 .

[0093] In step 33 of this implementation, assuming that the extracted rows are from the 3rd to the 98th row, the two-dimensional array M n-max-1 Identified as:

[0094] [F max-3-1 ,F max-3-2 ,F max-3-3 ,3;

[0095] F max-4-1 ,F max-4-2 ,F max-4-3 ,4;

[0096] ……;

[0097] F max-98-1 ,F max-98-2 ,F max-98-3 ,98];

[0098] Step 34, from the two-dimensional array M n-max-1 Select the largest common factor F in the second column max-2 , and the number of rows with the largest common number is N max-2 ;

[0099] Step 35, from the two-dimensional array M n-max-1 Extract the second number in each row equal to F max-2 All rows form a new 4-column N max-2 Two-dimensional array M of rows n-max-2 ;

[0100] In step 35 of this implementation, assuming that the extracted rows are from the 4th to the 85th, the two-dimensional array M n-max-2 Identified as:

[0101] [F max-6-1 ,F max-6-2 ,F max-6-3 ,6;

[0102] F max-7-1 ,F max-7-2 ,F max-7-3 ,7;

[0103] ……;

[0104] F max-90-1 ,F max-90-2 ,F max-90-3 ,90];

[0105] Step 36, from the two-dimensional array M n-max-2 Select the largest common factor F in the third column max-3 , and the number of rows with the largest common number is N max-3 ;

[0106] Step 37, from the two-dimensional array M n-max-2 Extract the third number in each row equal to F max-3 All rows form a new 4-column N max-3 Two-dimensional array M of rows n-max-3 .

[0107] In step 37 of this implementation, assuming that the extracted rows are from the 3rd to the 85th row, the two-dimensional array M n-max-3 Identified as:

[0108] [F max-8-1 ,F max-8-2 ,F max-8-3 ,8;

[0109] F max-7-1 ,F max-7-2 ,F max-7-3 ,7;

[0110] ……;

[0111] F max-90-1 ,F max-90-2 ,F max-90-3 ,90];

[0112] In one implementation of the embodiment of the present invention, the above step 4 includes the following situations:

[0113] If N max-1 <50%M n, then the test data is confirmed to be invalid;

[0114] If 50% M n ≤N max-1 ≤70%M n , then press M n-max-1 The fourth number (i.e. the nth circle) in each row is used to extract the corresponding circle and average it for the whole period. The averaged data is used for trend analysis and quantitative analysis with low confidence level.

[0115] If N max-1 >70%M n , and N max-2 ≤60%M n , then press M n-max-3 The fourth number in each row (i.e., the nth circle) is used to extract the corresponding circle for full-cycle averaging, and the test data of this round is considered valid and can be used for trend analysis and quantitative analysis with medium confidence.

[0116] If N max-1 >70%M n , and N max-3 ≤50%M n , then press M n-max-3 The fourth number in each row (i.e., the nth circle) is used to extract the corresponding circle for full-cycle averaging, and the test data of this round are considered valid and can be used for trend analysis and quantitative analysis with medium confidence.

[0117] If N max-1 >70%M n , and N max-2 >60%M n And N max-3 >50%M n , then press M n-max-3 The fourth number in each row (i.e., the nth circle) is used to extract the corresponding circle for full-cycle averaging, and it is considered that the test data of this round is completely valid and can be used for trend analysis and quantitative analysis with high confidence.

[0118] Based on the method for processing abnormal aerodynamic noise of a model rotor caused by external airflow disturbances provided in the above-mentioned embodiment of the present invention, an embodiment of the present invention also provides a device for processing abnormal aerodynamic noise of a model rotor caused by external airflow disturbances, including: a data acquisition and segmentation module, an extraction module, a statistical module, and a data processing and identification module.

[0119] The data collection and segmentation module is used to segment the collected rotor aerodynamic noise data by circles according to the characteristic that the rotor aerodynamic noise has main frequency stability in a stable state;

[0120] An extraction module is used to perform amplitude and frequency analysis on the obtained data of each circle, and extract the frequency values ​​corresponding to the first three maximum amplitudes in the amplitude spectrum of each circle of data;

[0121] The statistical module counts the frequency values ​​extracted from all circles to obtain the main frequency with the largest number of frequencies;

[0122] The data processing and identification module removes the data of the circles corresponding to the main frequency that is not the most common, averages the data of the remaining circles over a whole period, and identifies the confidence level of the noise data.

[0123] Based on the requirement of eliminating disturbance signals in rotor aerodynamic noise test, the embodiment of the present invention provides a method and device for processing model rotor abnormal aerodynamic noise caused by external airflow disturbance. According to the characteristic that rotor aerodynamic noise has main frequency stability in the stable state, the collected rotor aerodynamic noise data is divided into whole cycles, and then the amplitude and frequency analysis of each circle of data is performed, and the frequency values ​​corresponding to the first three maximum amplitudes in the amplitude spectrum of each circle of data are extracted. The frequency values ​​extracted from all circles are counted to obtain the main frequency with the largest number of frequencies, and the data of the circle corresponding to the non-largest number of main frequencies are eliminated. Then, the data of the remaining circles are averaged for the whole cycle, and the confidence level of the noise data is judged, so as to effectively eliminate the rotor abnormal aerodynamic noise caused by sudden gusts, and then ensure that effective rotor rotation noise test data can be extracted from the rotor abnormal aerodynamic noise caused by external airflow disturbance in an open environment. By adopting the technical solution provided by the embodiment of the present invention, the process of continuous change of the rotor shaft tilt angle can be accurately simulated in a set of equations, and the dynamic response of the tilt-rotor helicopter in the transition state can be solved therefrom.

Claims

1. A method for eliminating abnormal aerodynamic noise of a model rotor caused by external airflow disturbance, characterized in that: include: Step 1, according to the characteristic that the rotor aerodynamic noise in a stable state has a main frequency stability, the collected rotor aerodynamic noise data is divided into circles; Step 2, performing amplitude and frequency analysis on the obtained data of each circle, and extracting the frequency values ​​corresponding to the first three maximum amplitudes in the amplitude spectrum of each circle of data; Step 3, count the frequency values ​​extracted from all circles to obtain the main frequency with the largest number of frequencies; Step 4: Eliminate the data of the circles corresponding to the main frequency that is not the most common, average the data of the remaining circles over the entire period, and determine the confidence level of the noise data.

2. The method for eliminating abnormal aerodynamic noise of a model rotor caused by external airflow disturbance according to claim 1 is characterized in that: The step 1 comprises: Step 11, synchronously collecting noise data and rotor speed pulse data of each measuring point at a sampling frequency St; Step 12, according to the sampling frequency St and the rotor speed Nr, calculate the number of data points per rotor rotation Np = St / Nr; Step 13: Select the noise data of the measurement point farthest from the center line of the model rotor shaft from each noise measurement point, and divide M into the measurement points M and M in sequence with any pulse rising edge of the rotor speed Nr as the starting point. n The noise data sequence has Np data points in each circle, that is, M n 1-dimensional arrays, each of which has a length of Np.

3. The method for eliminating abnormal aerodynamic noise of a model rotor caused by external airflow disturbance according to claim 2 is characterized in that: In the step 11, The collected rotor speed pulse data is a signal that changes from a low level to a high level every time the rotor rotates one circle, the high level lasts for a period of time, and then returns to a low level. The duty cycle of the high level duration period of each circle ranges from 20% to 30%, St≥1024×Nr; where Nr is the rotor speed; Among them, the number of rising edge data points is 4 to 10, and the time length T of data collection satisfies: T = M0 × Nr; M0 is the number of turns, and the value range of M0 is 200≥M0≥130.

4. The method for eliminating abnormal aerodynamic noise of a model rotor caused by external airflow disturbance according to claim 2 is characterized in that: The method of performing amplitude and frequency analysis on each circle of data in step 2 includes: Step 21: M obtained by segmentation in step 1 n The 1-dimensional arrays are analyzed for auto-power spectrum in turn, so as to obtain the amplitude 1-dimensional array and frequency 1-dimensional array of the auto-power spectrum of each circle of data; Among them, the one-dimensional array of the amplitude of the power spectrum of the first cycle of data is marked as [A 1-1 , A 1-2 …、A 1-Np / 2 ] and the one-dimensional array of frequency values ​​is identified as [F 1-1 、F 1-2 …、F 1-Np / 2 ], No. M n The one-dimensional array of the amplitude of the power spectrum of the loop data is marked as [A Mn-1 , A Mn-2 …、A Mn-Np / 2 ] and the one-dimensional array of frequency values ​​is identified as [F Mn-1 、F Mn-2 …、F Mn-Np / 2 ].

5. The method for eliminating abnormal aerodynamic noise of a model rotor caused by external airflow disturbance according to claim 4 is characterized in that: The method of extracting the frequency values ​​corresponding to the first three maximum amplitudes in the amplitude spectrum of each circle of data in step 2 includes: Step 22, according to the one-dimensional array of amplitudes of the power spectrum of each circle of data obtained in step 21, extract three frequency values ​​corresponding to the first three maximum amplitudes in the one-dimensional array of amplitudes of the circle; Among them, the frequency values ​​corresponding to the first three maximum amplitudes in the one-dimensional array of the first circle amplitude are marked as [F max-1-1 ,F max-1-2 ,F max-1-3 ], No. M n The frequency values ​​corresponding to the first three largest amplitudes in the one-dimensional array of circle amplitudes are marked as [F max-Mn-1 ,F max-Mn-2 ,F max-Mn-3 ].

6. The method for eliminating abnormal aerodynamic noise of a model rotor caused by external airflow disturbance according to claim 5 is characterized in that: The step 3 comprises: According to the sorting of amplitude size, for the frequency data of all circles, first extract the maximum mode and number of rows of the frequency corresponding to the first amplitude value, and then extract the maximum mode and number of rows of the frequency corresponding to the second amplitude value based on the maximum mode row corresponding to the first amplitude value, and then extract the maximum mode and number of rows of the frequency corresponding to the third amplitude value based on the maximum mode row corresponding to the second amplitude value.

7. The method for eliminating abnormal aerodynamic noise of a model rotor caused by external airflow disturbance according to claim 5, characterized in that: The step 3 comprises: Step 31, according to the M obtained in step 22 n The frequency values ​​corresponding to the first three maximum amplitudes in each circle generate 4 columns M of the following form n Two-dimensional array M of rows all Identified as: [F max-1-1 ,F max-1-2 ,F max-1-3 ,1; F max-2-1 ,F max-2-2 ,F max-2-3 ,2; ……； F max-Mn-1 ,F max-Mn-2 ,F max-Mn-3 ,M n ]; Step 32, from the two-dimensional array M all Select the largest common factor F in the first column max-1 , and the number of rows with the largest common number is N max-1 ; Step 33, from the two-dimensional array M all Extract the first number in each row equal to F max-1 All rows form a new 4-column N max-1 Two-dimensional array M of rows n-max-1 ; Step 34, from the two-dimensional array M n-max-1 Select the largest common factor F in the second column max-2 , and the number of rows with the largest common number is N max-2 ; Step 35, from the two-dimensional array M n-max-1 Extract the second number in each row equal to F max-2 All rows form a new 4-column N max-2 Two-dimensional array M of rows n-max-2 ; Step 36, from the two-dimensional array M n-max-2 Select the largest common factor F in the third column max-3 , and the number of rows with the largest common number is N max-3 ; Step 37, from the two-dimensional array M n-max-2 Extract the third number in each row equal to F max-3 All rows form a new 4-column N max-3 Two-dimensional array M of rows n-max-3 .

8. The method for eliminating abnormal aerodynamic noise of a model rotor caused by external airflow disturbance according to claim 7 is characterized in that: The step 4 includes the following situations: If N max-1 <50%M n , then the test data is confirmed to be invalid; If 50% M n ≤N max-1 ≤70%M n , then press M n-max-1 The fourth number in each row is used to extract the corresponding circle and average the whole cycle. The averaged data is used for trend analysis and quantitative analysis with low confidence level. If N max-1 >70%M n , and N max-2 ≤60%M n , then press M n-max-3 The fourth number in each row is used to extract the corresponding circle for full-cycle averaging, and the test data of this round is considered valid. The averaged data is used for trend analysis and quantitative analysis with medium confidence. If N max-1 >70%M n , and N max-3 ≤50%M n , then press M n-max-3 The fourth number in each row is used to extract the corresponding circle for full-cycle averaging, and the test data of this round is considered valid. The averaged data is used for trend analysis and quantitative analysis with medium confidence. If N max-1 >70%M n , and N max-2 >60%M n And N max-3 >50%M n , then press M n-max-3 The fourth number in each row is used to extract the corresponding circle for full-cycle averaging, and the test data of this round is considered to be completely valid. The averaged data is used for trend analysis and quantitative analysis with high confidence.

9. A device for processing abnormal aerodynamic noise of a model rotor caused by external airflow disturbance, characterized in that: include: Data collection and segmentation module, extraction module, statistics module, data processing and identification module; The data collection and segmentation module is used to segment the collected rotor aerodynamic noise data by circles according to the characteristic that the rotor aerodynamic noise has a main frequency stability in a stable state; An extraction module is used to perform amplitude and frequency analysis on the obtained data of each circle, and extract the frequency values ​​corresponding to the first three maximum amplitudes in the amplitude spectrum of each circle of data; The statistical module counts the frequency values ​​extracted from all circles to obtain the main frequency with the largest number of frequencies; The data processing and identification module removes the data of the circles corresponding to the main frequency that is not the most common, averages the data of the remaining circles over a whole period, and identifies the confidence level of the noise data.

10. A computer-readable storage medium, characterized in that: include: Memory and processor; The memory is configured to store executable instructions; The processor is specifically configured to implement the method for processing abnormal aerodynamic noise of a model rotor caused by eliminating external airflow disturbances according to any one of claims 1 to 8 when executing the executable instructions stored in the memory.