CN101718582A - Tone testing method of wind power generator set - Google Patents

Abstract

A tone testing method for a wind power generating set, which is provided with a prepolarized microphone and a fan operating parameter sensor group. The pre-polarized microphone noise signal is output to the noise signal conditioning collector for processing, and output to the computer through the Wi-Fi wireless data transmission module; the fan operating parameter sensor group outputs the fan operating parameter signal to the fan operating parameter signal acquisition conditioner for processing, and Output to computer through USB data transfer module. The computer uses the virtual instrument software to judge the audio tone of the collected signal to obtain the tone audibility. The remarkable effects of the present invention are: real-time signal collection, high synchronization, high measurement accuracy, using order ratio analysis technology, eliminating frequency ambiguity, simulating human hearing and psychological characteristics, accurate and reasonable results, simple operation, and high cost performance. Tone analysis of a wind turbine at various full wind speeds.

Description

Wind turbine tone test method

technical field

The invention belongs to the field of wind power generator audio testing, and in particular relates to a multi-parameter testing instrument for testing and analyzing the tone of a wind power generating set, which is suitable for accurately testing and analyzing the tone at various wind speeds during the noise test of a wind power generating set, so as to Evaluate the fan pitch audibility, thus providing an objective basis for fan noise reduction and noise assessment.

Background technique

According to the technical requirements of IEC 61400-11 noise characteristics test of wind turbines, the acoustic noise test of wind turbines needs to be carried out under the specified geographical conditions at the same time for power, wind speed, wind direction, air temperature, air pressure, yaw angle, and pitch angle. , speed and A-weighted sound pressure level tests, and tone analysis at each full wind speed in order to evaluate the fan tone audibility, thus providing an objective basis for noise assessment and fan noise reduction. The pitch of wind turbine noise is caused by the interaction of meshing gears, aerodynamic instabilities with the surface of the rotor blades, and unsteady airflow acting on holes, cracks, or blunt trailing edges. Smooth signal. However, the existing wind turbine noise tone judgment adopts a narrow-band spectrum judgment method based on stationary signal analysis, which will bring frequency ambiguity to the judgment results. At the same time, conventional acoustic noise testers for wind turbines usually use split instruments such as multi-channel data recorders, sound level meters, 1/3 octave spectrum analyzers, and spectrum analyzers for measurement and analysis.

Disadvantages of the existing wind turbine tone test: low measurement accuracy, poor synchronization, serious frequency ambiguity, complicated operation, and high price.

Contents of the invention

The purpose of the present invention is to provide a tone testing method of a wind power generating set with high measurement accuracy, high synchronization, little influence of frequency ambiguity phenomenon, simple operation and low cost.

In order to achieve the above object, the technical scheme adopted in the present invention is as follows:

A method for testing the tone of a wind power generating set, the key of which is to carry out according to the following steps:

Step 1: Obtain wind speed signal, air temperature signal, air pressure signal, noise signal, rotational speed signal, electric power signal, installation height of anemometer, center height of wind power generator and surface roughness, which can be measured by instruments.

Step 2, to determine the wind speed, follow the steps below:

The first step is to convert the wind speed signal into a standard wind speed signal according to the air temperature signal, air pressure signal, anemometer installation height, wind generator center height and surface roughness, and the conversion of the standard wind speed is a known technology; wherein, The standard wind speed range will cover the full wind speed of 6m/s, 7m/s, 8m/s, 9m/s, 10m/s.

The second step is to define the initial value P=1 of the sequence number P of the rectified wind speed;

The third step is to select the Pth rectified wind speed in the standard wind speed signal;

The fourth step is to determine the wind speed V _p of the Pth wind speed;

In the international standard, the wind speed V ₁ =6m/s is usually selected as the first wind speed, and other initial wind speed values can be selected if necessary.

Step 3, determine the speed fitting curve, follow the steps below:

The first step is to determine the noise data segment group to be analyzed, the noise data segment group is composed of the first noise data segment and the second noise data segment: within the Pth full wind speed, select the closest wind speed _Vp A section of 1 minute noise signal of one section is as the first noise data section; Another section of 1 minute noise signal of the closest described wind speed size _V is selected as the second noise data section;

Second step, determine the noise sub-data segment group, this noise sub-data segment group is made up of the first noise sub-data segment group and the second noise sub-data segment group: the first noise data segment is divided into 6 noise sub-data Section, the time length of each noise sub-data segment is 10 seconds, as the first noise sub-data segment group; the second noise data segment is divided into 6 noise sub-data segments, the time length of each noise sub-data segment Be 10 seconds, as the second noise sub-data segment group;

The third step is to determine the rotational speed data segment group, which is composed of the first rotational speed data segment and the second rotational speed data segment: select the rotational speed signal collected simultaneously with the first noise data segment as the first rotational speed data segment, and select and The rotational speed signal collected simultaneously by the second noise data segment is used as the second rotational speed data segment;

The fourth step is to determine the rotational speed sub-data segment group, which is composed of the first rotational speed sub-data segment group and the second rotational speed sub-data segment group: the first rotational speed data segment is divided into 6 segments of rotational speed sub-data Section, the time length of each speed sub-data segment is 10 seconds, as the first speed sub-data segment group; the second speed data segment is divided into 6 speed sub-data segments, the time length of each speed sub-data segment 10 seconds, as the second speed sub-data segment group;

The fifth step is to determine the fitting curve R _k (t) of the kth sub-data segment of the rotational speed on the group of rotational speed data segments, and its expression is: R _k (t)=a _k t ² +b _k t+c _k , where t is the time, a _k , b _k , and c _k are the fitting curve coefficients; due to the low speed of the fan, it can be obtained by looking for the relationship between R _k (t) and t on the speed sub-data segment Segmented fitting curve coefficients to achieve high-precision fitting of fan speed curves.

Step 4, determine the time scale of the phase identification, follow the steps below:

The first step, determine the sampling frequency f s of described fitting curve R _k (t) _{, k} , its expression is: f _{s, k} ≥ 2O _{max, k} , O _{max, k} is described fitting curve R _k (t) the largest order component. The equiangular sampling interval satisfies the sampling theorem, which makes the equiangular sampling of the phase detection time scale to the fitting curve more reasonable.

The second step is to determine the equiangular sampling interval Δθ _k of the fitting curve R _k (t), its expression is: Δθ _k =1/f _s,k ;

The 3rd step, determine the integral equation of described fitting curve R _k (t), its expression is:

Wherein, n is the time scale sequence number, T _{0, k} is the initial moment of described fitting curve R _k (t), and T _{-1, k} is the terminal moment of this fitting curve R _k (t);

确定所述第k段噪声子数据段的鉴相时标T_n，k。为方便获得准确的鉴相时标T_n，k，可取f_s，k＝2O_max，k。The 4th step, _{obtain formula}

Determine the phase detection time scale T _n,k of the kth noise sub-data segment. In order to obtain an accurate phase detection time scale T _{n, k} conveniently, f _{s, k} = 2O _{max, k} may be taken.

其中，t_0，k为所述第k段噪声子数据段中时间坐标小于等于T_n，k的第一个点，t_1，k为t_0，k的下一个点；采用鉴相时标对待分析的一分钟长度噪声信号进行线性插值，实现等角度重采样，可获得准平稳的噪声信号。Step 5, determine the quasi-stationary noise signal x(Tn _{, k} ) of the kth section noise sub-data section, its expression is:

Wherein, t _{0, k} is the first point of time coordinates less than or equal to T _{n, k} in the described k section noise sub-data segment, and t _{1, k} is t _{0, the next point of k} ; Adopt phase identification time scale Linear interpolation is performed on the one-minute noise signal to be analyzed to realize equiangular resampling, and a quasi-stationary noise signal can be obtained.

Step 6: Carry out A-weighting on the quasi-stationary noise signal x(T _{n, k} ), simulate the hearing of the human ear, add a Hanning window to reduce the energy leakage of the noise signal, perform FFT transformation, and obtain the noise sub-data of the kth segment The A-weighted narrowband sound pressure spectrum level L _A-NS,k of the section;

Step seven, judge the tone, follow the steps below:

The first step is to determine the critical frequency band S _k at the A-weighted narrowband sound pressure spectrum level L _A-NS,k , the expression of which is: Among them _{, f c and k} are the center frequencies of the kth noise sub-data segment; according to the auditory masking effect of the human ear, the critical frequency band is selected as the critical analysis range for subsequent judgment of whether the tone can be perceived by the human ear.

其中，M为所述临界频带S_k内幅值最小的70％根谱线的总数，L_m，k为该临界频带S_k内第m个幅值最小谱线的噪声频谱能量；The second step is to select 70% of the spectral lines with the smallest amplitude in the critical frequency band S _k to determine the criterion level L _{70%, k} , whose expression is:

Wherein, M is the total number of 70% spectral lines with the smallest amplitude in the critical frequency band S _k , and L _{m, k} is the noise spectrum energy of the mth smallest amplitude spectral line in the critical frequency band S _k ;

The third step is to select the spectral line whose amplitude is less than L _{70%, k} +6dB in the critical frequency band _Sk , and determine it as the masking noise in the critical frequency band _Sk ;

其中，H为所述掩蔽噪声的谱线总数，L_h，k为所述临界频带S_k内第h个掩蔽噪声的噪声频谱能量；The fourth step is to determine the average masking noise level L _pn,avg,k in the critical frequency band S _k , the expression of which is:

Wherein, H is the total number of spectral lines of the masking noise, L _{h, k} is the noise spectrum energy of the hth masking noise in the critical frequency band S _k ;

The fifth step, in the critical frequency band S _k , select the spectral line whose noise spectrum energy is greater than _{Lpn, avg, k} +6dB as the tone spectral line in the critical frequency band S _k , and the tone may appear on the tone spectral line middle.

其中，G为所述音调谱线的总数，L_g，k为所述临界频带S_k内第g个音调谱线的噪声频谱能量；The sixth step is to sum the tone spectral lines in decibels to determine the tone level L _pt0,k of the spectral lines in the critical frequency band S _k , the expression of which is:

Wherein, G is the total number of the tone spectrum lines, Lg _{, k} is the noise spectrum energy of the gth tone spectrum line in the critical frequency band _Sk ;

The seventh step is to judge whether the number i of adjacent tone spectral lines in the critical frequency band S _k is greater than 1: if i=1, then this tone spectral line is used as the tone in the critical frequency band S _k , and the tone level L _pt,k =L _pt0,k ; if i>1, the spectral line with the largest amplitude among the adjacent tone spectral lines is used as the tone in the critical frequency band S _k , and the tone level L _pt,k =L _{pt0 , k} /1.5;

When the tone contains more than two adjacent spectral lines, considering the spread effect of the Hanning window to increase the equivalent continuous bandwidth of the FFT, divide the tone level of the spectral line by 1.5 to make the tone test results more accurate and the test method more accurate Reasonable. In addition, spectral lines other than masking noise and tones in the critical frequency band will not be used for subsequent analysis.

Step 8, correct the background noise, follow the steps below:

The first step is to determine the equivalent continuous sound pressure level L n, _k of the background noise in the critical frequency band S _k when the fan is stopped according to the electric power signal;

The second step is to determine the modified masking noise level L _s,pn,avg,k within the critical frequency band S _k , the expression of which is:

${\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; pn</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; avg</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 101</span>}\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; 10</span>}}^{\mathrm{<span\; class="notranslate">\; 0.1</span>}{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; pn</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; avg</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 10</span>}}^{\mathrm{<span\; class="notranslate">\; 0.1</span>}{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span>$

To ensure the validity of the analysis results, it must be confirmed that the tonal components do not come from background noise. After the background noise is corrected, the obtained corrected masking noise level can more accurately express the equivalent continuous sound pressure level of the fan when it is running, that is, the fan noise without background noise.

Step 9, determining the masking noise level L _pn,k in the critical frequency band S _k and the final tone level ΔL _tn,k of the kth noise sub-data segment in the Pth full wind speed, according to the following steps:

The first step is to determine the effective noise bandwidth Z, whose expression is: Z=1.5×R, where R is the frequency resolution; the effective noise bandwidth Z corrects the influence of using the Hanning window, making the test results more accurate and reasonable.

The second step is to determine the masking noise level L _pn,k , the expression of which is:

The third step is to determine the final pitch level ΔL _{tn, k} , its expression is: ΔL _{tn, k} = L _{pt, k} - L _{pn, k} ;

Step ten, determine the audibility of the tone, follow the steps below:

The first step is to determine the audibility ΔL _t of the Pth full wind speed, the expression of which is:

$\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 101</span>}\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 12</span>}}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 12</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; 10</span>}}^{\mathrm{<span\; class="notranslate">\; 0.1</span>}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; tn</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span>$

其中f_tone为音调频率值；The second step is to determine the frequency-based audibility criterion L _a , which is used to compensate the human ear's response to tones of different frequency components, and its expression is:

Where f _tone is the tone frequency value;

The third step is to determine the tone audibility ΔL _a , its expression is: ΔL _a = ΔL _t -L _a ;

Step 11, judging whether the number P of the rectified wind speed is less than P', if P is less than P', then add 1 to P, and return to the step of selecting the Pth rectified wind speed in the standard wind speed signal, that is, the first step of step 2 Three steps; if P is equal to or greater than P', the analysis ends; among them, P' is the serial number of the end rectification wind speed. In international standards, the wind speed V _P ′=10m/s is usually selected as the last wind speed to be analyzed, that is, the wind speeds of 6m/s, 7m/s, 8m/s, 9m/s, 10m/s rectified wind speed for analysis, if necessary, more rectified wind speeds can be selected. At each selected full wind speed, all the above analysis steps are repeated until the tones at each selected full wind speed are analyzed.

In the step of obtaining wind speed signal, air temperature signal, air pressure signal, noise signal, speed signal, electric power signal, anemometer installation height, wind power generator center height and surface roughness, that is, in step one, use a prepolarized microphone Acquire the noise signal, process the noise signal with a noise signal conditioning collector, acquire the wind speed signal, air temperature signal, air pressure signal, rotational speed signal and electric power signal with the fan operating parameter sensor group, and process it with the fan operating parameter signal acquisition conditioner The wind speed signal, air temperature signal, air pressure signal, rotational speed signal and electric power signal.

Wherein, the prepolarized microphone is used to acquire noise signals. Its signal output end is connected to the signal acquisition end of the noise signal conditioning collector, and the noise signal conditioning collector performs anti-aliasing filtering on the collected noise data to improve the signal-to-noise ratio during analysis, and at the same time improve the accuracy of the phase detection time scale. Accuracy when performing linear interpolation on noisy data. The Wi-Fi wireless data output end of the noise signal conditioning collector is wirelessly connected with the Wi-Fi wireless data input end of the computer, and the noise signal data after A/D conversion is provided to the computer for processing through wireless transmission. The signal output end of the fan operation parameter sensor group is connected to the signal acquisition end group of the fan operation parameter signal acquisition conditioner, and the USB output end of the fan operation parameter signal acquisition conditioner is connected to the USB input end of the computer, and the collected The received data is converted into a voltage signal and transmitted to the computer for processing through the USB interface.

The noise signal conditioning collector is provided with a built-in piezoelectric ICP excitation constant current source, an AC coupler, an anti-aliasing filter, a first A/D converter and a Wi-Fi wireless data transmission module, and the built-in piezoelectric ICP excitation The output terminal of the constant current source is connected to the power supply terminal of the prepolarized microphone, and is used for providing constant current excitation to the prepolarized microphone. The signal output end of the prepolarized microphone is connected to the signal acquisition end of the AC coupler, the signal output end of the AC coupler is connected to the signal input end of the anti-aliasing filter, and the signal output end of the anti-aliasing filter is connected to The signal input end of the first A/D converter, the signal output end of the first A/D converter is connected to the signal input end of the Wi-Fi wireless data transmission module, the Wi-Fi wireless data transmission module The signal output end is wirelessly connected with the Wi-Fi wireless data input end of the computer; the noise signal conditioning collector is powered by a rechargeable battery pack.

The prepolarized microphone sends the collected noise signal to the noise signal conditioning collector for signal conditioning and A/D conversion, and then transmits the digital signal to the computer through the Wi-Fi wireless communication module to realize real-time collection of the noise signal.

The fan operation parameter signal acquisition conditioner is provided with a multiplexer, an amplification filter, a second A/D converter and a USB data transmission module, and the signal output end of the fan operation parameter sensor group is connected to the multiplexer The signal acquisition end group of the device, the signal output end of the multiplexer is connected to the signal input end of the amplification filter, and the signal output end of the amplification filter is connected to the signal input end of the second A/D converter, The signal output end of the second A/D converter is connected to the signal input end of the USB data transmission module, and the signal output end of the USB data transmission module is connected to the USB input end of the computer.

The fan operating parameter sensor group converts the fan operating parameters into a voltage signal, and the fan operating parameter signal acquisition conditioner performs conditioning and A/D conversion on the voltage signal, and then transmits the digital signal to the computer in real time through the USB data transmission module to realize fan monitoring. Real-time collection of operating parameters.

The fan operating parameter sensor group is composed of an electric power sensor, a wind speed sensor, an air temperature sensor, an air pressure sensor and a rotational speed sensor, which are respectively used to obtain electric power signals, wind speed signals, air temperature signals, air pressure signals and rotational speed signals. The signal output end of the electric power sensor is connected to the electric power signal acquisition end of the fan operation parameter signal acquisition conditioner, the signal output end of the wind speed sensor is connected to the wind speed signal acquisition end of the fan operation parameter signal acquisition conditioner, and the air temperature The signal output end of the sensor is connected to the air temperature signal acquisition end of the fan operation parameter signal acquisition conditioner, the signal output end of the air pressure sensor is connected to the air pressure signal acquisition end of the fan operation parameter signal acquisition conditioner, and the signal output of the speed sensor is The end is connected to the speed signal acquisition end of the fan operation parameter signal acquisition conditioner.

The test software of the computer is a virtual instrument. The virtual instrument software in the computer realizes the simultaneous acquisition of the noise signal and the fan operating parameter signal through modularization and hardware triggering, so as to realize the tone analysis at the standard wind speed. This method can complete functions such as signal communication, signal preprocessing, non-acoustic signal analysis, tone analysis, data display, data management, report printing and output, and network communication, and can perform off-line analysis and pre-simulation without hardware.

The remarkable effects of the present invention are: real-time signal collection, high synchronization, high measurement accuracy, using order ratio analysis technology, eliminating frequency ambiguity, simulating human hearing and psychological characteristics, accurate and reasonable results, simple operation, and high cost performance. Tone analysis of a wind turbine at various full wind speeds.

Description of drawings

Fig. 1 is a block diagram of the device structure of the present invention.

Fig. 2 is the main flowchart of the present invention;

Fig. 3 is the flow chart of determining rotational speed fitting curve;

Fig. 4 is the flow chart of determining phase identification time scale;

Fig. 5 is the flowchart of judging tone;

Fig. 6 is the flowchart of correcting background noise;

Fig. 7 is the flowchart of determining masking noise level and final tone level;

FIG. 8 is a flowchart for determining tone audibility.

Detailed ways

The present invention will be described in further detail below in conjunction with the accompanying drawings and specific embodiments.

As shown in Figure 1, a method for testing the tone of a wind power generating set firstly needs to obtain the wind speed signal, the air temperature signal, the air pressure signal, the noise signal, the rotational speed signal and the electric power signal, adopt the prepolarized microphone 1 to obtain the described noise signal, and use the noise signal The signal conditioning collector 2 processes the noise signal, acquires the wind speed signal, air temperature signal, air pressure signal, rotational speed signal and electric power signal with the fan operating parameter sensor group 3, and processes the wind speed signal, air temperature signal with the fan operating parameter signal acquisition conditioner 4 signal, air pressure signal, speed signal and electric power signal;

The output end of the built-in piezoelectric ICP excitation constant current source 6 of the noise signal conditioning collector 2 is connected to the power supply end of the prepolarized microphone 1, and the signal output end of the prepolarized microphone 1 is connected to the signal acquisition end of the AC coupler 7. The signal output end of the AC coupler 7 is connected to the signal input end of the anti-aliasing filter 8, and the signal output end of the anti-aliasing filter 8 is connected to the signal input end of the first A/D converter 9, and the first A/D conversion The signal output end of device 9 connects the signal input end of Wi-Fi wireless data transmission module 10, and the signal output end of this Wi-Fi wireless data transmission module 10 is wirelessly connected with the Wi-Fi wireless data input end of computer 5; Noise signal conditioning The collector 2 is powered by a rechargeable battery pack.

The signal output end of the fan operation parameter sensor group 3 is connected to the signal acquisition end group of the multiplexer 11 of the fan operation parameter signal acquisition conditioner 4, and the signal output end of the multiplexer 11 is connected to the signal input end of the amplification filter 12 , the signal output end of the amplification filter 12 is connected to the signal input end of the second A/D converter 13, and the signal output end of the second A/D converter 13 is connected to the signal input end of the USB data transmission module 14, the USB The signal output end of the data transmission module 14 is connected to the USB input end of the computer 5 .

As shown in Figure 2, a method for testing the tone of a wind power generating set is carried out according to the following steps:

Obtain wind speed signal, air temperature signal, air pressure signal, noise signal, speed signal, electric power signal, anemometer installation height, wind turbine center height and surface roughness;

Converting the wind speed signal into a standard wind speed signal according to the air temperature signal, air pressure signal, installation height of the anemometer, center height of the wind generator and surface roughness;

Define the initial value P=1 of the sequence number P of the rectified wind speed;

Selecting the Pth rectified wind speed in the standard wind speed signal;

Determine the wind speed V _p of the Pth full wind speed, where the initial analysis wind speed V ₁ =6m/s.

Determine the speed fitting curve;

Determine phase identification time scale;

Determine the quasi-stationary noise signal x(T _{n, k} ) of the kth section noise sub-data segment, its expression is:

Wherein, t _{0, k} is the time coordinate less than or equal to T _{n in the described k section noise sub-data segment, the first point of k} , t _{1, k} is t _{0, the next point of k} ;

Perform A-weighting, Hanning window, and FFT transformation on the quasi-stationary noise signal x(T _{n, k} ) to obtain the A-weighted narrowband sound pressure spectrum level L _{A-NS, k} of the noise sub-data segment of the kth segment ;

judge tone;

Correct background noise;

Determining the masking noise level L _{pn , k} in the critical frequency band S k and the final tone level ΔL _{tn, k} of the kth noise sub-data segment in the Pth full wind speed;

Determine tonal audibility;

Judging whether the number P of the rectified wind speed is less than P', if P is less than P', that is, V _p <10m/s, then add 1 to P, and return to the step of selecting the Pth rectified wind speed in the standard wind speed signal; if If P is equal to or greater than P', that is, V _p ≥ 10m/s, then the analysis ends; where, V _P '=10m/s, is the wind speed at which the analysis ends.

As shown in Figure 3, to determine the speed fitting curve, follow the steps below:

Determine the noise data segment group to be analyzed, the noise data segment group is composed of the first noise data segment and the second noise data segment: within the Pth full wind speed, select a section of 1 minute closest to the wind speed V _p The noise signal is used as the first noise data segment; another section of 1 minute noise signal closest to the wind speed _V is selected as the second noise data segment;

Determine the noise sub-data segment group, this noise sub-data segment group is made up of the first noise sub-data segment group and the second noise sub-data segment group: the first noise data segment is divided into 6 noise sub-data segments, each segment The time length of the noise sub-data segment is 10 seconds, as the first noise sub-data segment group; the second noise data segment is divided into 6 noise sub-data segments, and the time length of each noise sub-data segment is 10 seconds, as the second noise sub-segment group;

Determine the rotational speed data segment group, the rotational speed data segment group is composed of the first rotational speed data segment and the second rotational speed data segment: select the rotational speed signal collected simultaneously with the first noise data segment as the first rotational speed data segment, and select the second rotational speed data segment as the first rotational speed data segment The rotational speed signal collected simultaneously by the segment is used as the second rotational speed data segment;

Determine the speed sub-data segment group, the speed sub-data segment group is composed of the first speed sub-data segment group and the second speed sub-data segment group: the first speed data segment is divided into 6 speed sub-data segments, each segment The time length of the speed sub-data segment is 10 seconds, as the first speed sub-data segment group; the second speed data segment is divided into 6 sections of speed sub-data segments, and the time length of each speed sub-data segment is 10 seconds, as the second rotational speed sub-segment group;

On the rotational speed data segment group, determine the fitting curve R _k (t) of the kth rotational speed sub-data segment, the expression of which is: R _k (t)=a _k t ² +b _k t+c _k , wherein, t is time, a _k , b _k , c _k are polynomial coefficients.

As shown in Figure 4, to determine the phase identification time scale, follow the steps below:

Determine the sampling frequency f s of described fitting curve R _k (t) _{, k} , its expression is: f _{s, k} =2O _{max, k} , O _{max, k} is the value of described fitting curve R _k (t) largest order component;

Determine the equiangular sampling interval Δθ _k of the fitting curve R _k (t), its expression is: Δθ _k =1/f _{s, k} ;

Determine the integral equation of the fitting curve R _k (t), its expression is:

其中，n为时标序号，T_0，k为所述拟合曲线R_k(t)的初始时刻，T_-1，k为该拟合曲线R_k(t)的终点时刻；

Wherein, n is the time scale sequence number, T _{0, k} is the initial moment of described fitting curve R _k (t), and T _{-1, k} is the terminal moment of this fitting curve R _k (t);

确定所述第k段噪声子数据段的鉴相时标T_n，k；According to the integral equation of described equiangular sampling interval Δθ _k and fitting curve R _k (t), obtain formula

Determining the phase detection time scale Tn _,k of the kth noise sub-data segment;

As shown in Figure 5, to judge the tone, follow the steps below:

其中f_c，k为第k段噪声子数据段的中心频率；On the A-weighted narrowband sound pressure spectrum level L _A-NS,k , determine the critical frequency band S _k , its expression is:

Wherein f _{c, k} is the center frequency of the kth segment noise sub-data segment;

其中，M为所述临界频带S_k内幅值最小的70％根谱线总数，L_m，k为该临界频带S_k内第m个幅值最小谱线的噪声频谱能量；Select 70% of the spectral lines with the smallest amplitude in the critical frequency band S _k to determine the criterion level L _{70%, k} , whose expression is:

Wherein, M is the total number of 70% spectral lines with the smallest amplitude in the critical frequency band S _k , and L _{m, k} is the noise spectrum energy of the mth smallest amplitude spectral line in the critical frequency band S _k ;

Selecting the spectral line whose amplitude is less than L _{70%, k} +6dB in the critical frequency band _Sk is determined as the masking noise in the critical frequency band _Sk ;

其中，H为所述掩蔽噪声的谱线总数，L_h，k为所述临界频带S_k内第h个掩蔽噪声的噪声频谱能量；Determine the average masking noise level L _pn,avg,k within the critical frequency band S _k , its expression is:

Wherein, H is the total number of spectral lines of the masking noise, L _{h, k} is the noise spectrum energy of the hth masking noise in the critical frequency band S _k ;

In the critical frequency band S _k , select the spectral line whose noise spectrum energy is greater than _{Lpn, avg, k} +6dB as the tone spectral line in the critical frequency band S _k ;

其中，G为所述音调谱线的总数，L_g，k为所述临界频带S_k内第g个音调谱线的噪声频谱能量；Determine the spectral line tone level L pt0 in the critical frequency band S _{k , the expression of k} is:

Wherein, G is the total number of the tone spectrum lines, Lg _{, k} is the noise spectrum energy of the gth tone spectrum line in the critical frequency band _Sk ;

Judging whether the number i of adjacent tone spectral lines in the critical frequency band S _k is greater than 1: if i=1, then the tone spectral lines are used as the tone in the critical frequency band S _k , and the tone level L _{pt, k} = L _{pt0, k} ; if i>1, the spectral line with the largest amplitude among the adjacent tone spectral lines is used as the tone in the critical frequency band S _k , and the tone level L _{pt, k} =L _{pt0, k} /1.5 ;

As shown in Figure 6, to correct the background noise, follow the steps below:

According to the electric power signal, determine the equivalent continuous sound pressure level L _n,k of the background noise in the critical frequency band S _k when the fan is stopped;

Determine the modified masking noise level L _s,pn,avg,k within the critical frequency band S _k , the expression of which is:

${\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; pn</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; avg</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 101</span>}\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; 10</span>}}^{\mathrm{<span\; class="notranslate">\; 0.1</span>}{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; pn</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; avg</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 10</span>}}^{\mathrm{<span\; class="notranslate">\; 0.1</span>}{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span>$

As shown in Figure 7, to determine the masking noise level L _{pn ,k} in the critical frequency band S k and the final tone level ΔL _tn,k of the kth noise sub-data segment in the Pth full wind speed, follow the following steps conduct:

Determine the effective noise bandwidth Z, its expression is: Z=1.5×R, where R is the frequency resolution;

Determine the masking noise level L _pn,k , its expression is:

Determine the final pitch level ΔL _tn,k , the expression of which is: ΔL _tn,k =L _pt,k -L _pn,k ;

As shown in Figure 8, to determine the tone audibility, follow the steps below:

Determine the audibility ΔL _t of the Pth full wind speed, its expression is:

$\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 101</span>}\mathrm{<span\; class="notranslate">\; g</span>}<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 12</span>}}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 12</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; 10</span>}}^{\mathrm{<span\; class="notranslate">\; 0.1</span>}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; tn</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span>$

Determine the frequency-based audibility criterion L _a , whose expression is:

其中f_tone为音调频率值；

Where f _tone is the tone frequency value;

To determine the tone audibility ΔL _a , its expression is: ΔL _a =ΔL _t -L _a .

Its working conditions are as follows: the pre-polarization microphone 1 outputs the collected noise signal to the noise signal conditioning collector 2, and after being converted and processed by the noise signal conditioning collector 2, it is output to the computer 5 through the Wi-Fi wireless data transmission module 10. Simultaneously, the fan operation parameter sensor group 3 outputs the fan operation parameter signal collected to the fan operation parameter signal acquisition conditioner 4, and after the fan operation parameter signal acquisition conditioner 4 is converted and processed, it is output to the computer 5 through the USB data transmission module 14 . The computer 5 fits the rotation speed signal, determines the phase detection time scale of equiangular sampling, resamples the noise signal at an equal angle to obtain a quasi-stationary noise signal, and judges and evaluates the noise and tone, corrects the background noise, and obtains the tone audibility.

Claims (6)

1. A method for testing the tone of a wind power generating set, characterized in that, proceed according to the following steps: Step 1, obtaining wind speed signal, air temperature signal, air pressure signal, noise signal, rotational speed signal, electric power signal, anemometer installation height, wind turbine center height and surface roughness; Step 2, to determine the wind speed, follow the steps below: The first step is to convert the wind speed signal into a standard wind speed signal according to the air temperature signal, air pressure signal, anemometer installation height, wind turbine center height and surface roughness; The second step is to define the initial value P=1 of the sequence number P of the rectified wind speed; The third step is to select the Pth rectified wind speed in the standard wind speed signal; The fourth step is to determine the wind speed V _p of the Pth wind speed; Step 3, determine the speed fitting curve, follow the steps below: The first step is to determine the noise data segment group to be analyzed, the noise data segment group is composed of the first noise data segment and the second noise data segment: within the Pth full wind speed, select the closest wind speed _Vp A section of 1 minute noise signal of one section is as the first noise data section; Another section of 1 minute noise signal of the closest described wind speed size _V is selected as the second noise data section; Second step, determine the noise sub-data segment group, this noise sub-data segment group is made up of the first noise sub-data segment group and the second noise sub-data segment group: the first noise data segment is divided into 6 noise sub-data Section, the time length of each noise sub-data segment is 10 seconds, as the first noise sub-data segment group; the second noise data segment is divided into 6 noise sub-data segments, the time length of each noise sub-data segment Be 10 seconds, as the second noise sub-data segment group; The third step is to determine the rotational speed data segment group, which is composed of the first rotational speed data segment and the second rotational speed data segment: select the rotational speed signal collected simultaneously with the first noise data segment as the first rotational speed data segment, and select and The rotational speed signal collected simultaneously by the second noise data segment is used as the second rotational speed data segment; The fourth step is to determine the rotational speed sub-data segment group, which is composed of the first rotational speed sub-data segment group and the second rotational speed sub-data segment group: the first rotational speed data segment is divided into 6 segments of rotational speed sub-data Section, the time length of each speed sub-data segment is 10 seconds, as the first speed sub-data segment group; the second speed data segment is divided into 6 speed sub-data segments, the time length of each speed sub-data segment 10 seconds, as the second speed sub-data segment group; The fifth step is to determine the fitting curve R _k (t) of the kth sub-data segment of the rotational speed on the group of rotational speed data segments, and its expression is: R _k (t)=a _k t ² +b _k t+c _k , where t is time, a _k , b _k , c _k are polynomial coefficients; Step 4, determine the time scale of the phase identification, follow the steps below: The first step, determine the sampling frequency f s of described fitting curve R _k (t) _{, k} , its expression is: f _{s, k} ≥ 2O _{max, k} , O _{max, k} is described fitting curve R _k the largest order component of (t); The second step is to determine the equiangular sampling interval Δθ _k of the fitting curve R _k (t), its expression is: Δθ _k =1/f _s,k ; The 3rd step, determine the integral equation of described fitting curve R _k (t), its expression is:

Wherein, n is the time scale sequence number, T _{0, k} is the initial moment of described fitting curve R _k (t), and T _{-1, k} is the terminal moment of this fitting curve R _k (t); The 4th step _, obtain _formula

Determining the phase detection time scale Tn _,k of the kth noise sub-data segment; Step 5, determine the quasi-stationary noise signal x(Tn _{, k} ) of the kth section noise sub-data section, its expression is: Wherein, t _{0, k} is the time coordinate less than or equal to T _{n in the described k section noise sub-data segment, the first point of k} , t _{1, k} is t _{0, the next point of k} ; Step 6: Perform A-weighting, Hanning window, and FFT transformation on the quasi-stationary noise signal x(T _{n, k} ) to obtain the A-weighted narrowband sound pressure spectrum level L _A- of the kth noise sub-data segment _NS,k ; Step seven, judge the tone, follow the steps below: The first step is to determine the critical frequency band S _k at the A-weighted narrowband sound pressure spectrum level L _A-NS,k , the expression of which is:

Wherein f _{c, k} is the center frequency of the kth segment noise sub-data segment; The second step is to select 70% of the spectral lines with the smallest amplitude in the critical frequency band S _k to determine the criterion level L _{70%, k} , whose expression is:

Wherein, M is the total number of 70% spectral lines with the smallest amplitude in the critical frequency band S _k , and L _{m, k} is the noise spectrum energy of the mth smallest amplitude spectral line in the critical frequency band S _k ; The third step is to select the spectral line whose amplitude is less than L _{70%, k} +6dB in the critical frequency band _Sk , and determine it as the masking noise in the critical frequency band _Sk ; The fourth step is to determine the average masking noise level L _pn,avg,k in the critical frequency band S _k , the expression of which is: Wherein, H is the total number of spectral lines of the masking noise, L _{h, k} is the noise spectrum energy of the hth masking noise in the critical frequency band S _k ; The fifth step, in the critical frequency band S _k , select the spectral line whose noise spectrum energy is greater than _{Lpn, avg, k} +6dB as the tone spectral line in the critical frequency band S _k ; The sixth step is to determine the spectral line tone level L _pt0,k in the critical frequency band S _k , the expression of which is:

Wherein, G is the total number of the tone spectrum lines, Lg _{, k} is the noise spectrum energy of the gth tone spectrum line in the critical frequency band _Sk ; The seventh step is to judge whether the number i of adjacent tone spectral lines in the critical frequency band S _k is greater than 1: if i=1, then this tone spectral line is used as the tone in the critical frequency band S _k , and the tone level L _pt,k =L _pt0,k ; if i>1, the spectral line with the largest amplitude among the adjacent tone spectral lines is used as the tone in the critical frequency band S _k , and the tone level L _pt,k =L _{pt0 , k} /1.5; Step 8, correct the background noise, follow the steps below: The first step is to determine the equivalent continuous sound pressure level Ln,k of the background noise in the critical frequency band Sk when the fan is stopped according to the electric power signal; The second step is to determine the modified masking noise level L _s,pn,avg,k within the critical frequency band S _k , the expression of which is: ${\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; pn</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; avg</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 10</span>}\mathrm{<span\; class="notranslate">\; lg</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; 10</span>}}^{{\mathrm{<span\; class="notranslate">\; 0.1</span>}\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; pn</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; avg</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; 10</span>}}^{\mathrm{<span\; class="notranslate">\; 0.1</span>}{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span>$ Step 9, determining the masking noise level L _pn,k in the critical frequency band S _k and the final tone level ΔL _tn,k of the kth noise sub-data segment in the Pth full wind speed, according to the following steps: The first step is to determine the effective noise bandwidth Z, its expression is: Z=1.5×R, where R is the frequency resolution; The second step is to determine the masking noise level L _pn,k , the expression of which is:

The third step is to determine the final pitch level ΔL _{tn, k} , its expression is: ΔL _{tn, k} = L _{pt, k} - L _{pn, k} ; Step ten, determine the audibility of the tone, follow the steps below: The first step is to determine the audibility ΔL _t of the Pth full wind speed, the expression of which is: $\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 10</span>}\mathrm{<span\; class="notranslate">\; lg</span>}<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 12</span>}}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 12</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; 10</span>}}^{\mathrm{<span\; class="notranslate">\; 0.1</span>}\mathrm{<span\; class="notranslate">\; \&Delta;</span>}{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; tn</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; k</span>}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ;</span>$ The second step is to determine the frequency-based audibility criterion L _a , whose expression is:

Where f _tone is the tone frequency value; The third step is to determine the tone audibility ΔL _a , its expression is: ΔL _a = ΔL _t -L _a ; Step eleven, judging whether the number P of the rectified wind speed is less than P', if P is less than P', add 1 to P, and return to the step of selecting the Pth rectified wind speed in the standard wind speed signal; if P is equal to or greater than P', the analysis ends; among them, P' is the sequence number of the ending wind speed adjustment. 2. The method for testing the tone of a wind power generating set according to claim 1, characterized in that: in said acquisition of wind speed signal, air temperature signal, air pressure signal, noise signal, rotational speed signal, electric power signal, anemometer installation height, wind generator In the steps of center height and surface roughness, the noise signal is acquired with a prepolarized microphone (1), processed with a noise signal conditioning collector (2), and the noise signal is acquired with a fan operating parameter sensor group (3). The wind speed signal, air temperature signal, air pressure signal, speed signal and electric power signal are processed by the fan operation parameter signal acquisition conditioner (4) to process the wind speed signal, air temperature signal, air pressure signal, speed signal and electric power signal; wherein, the prepolarization The signal output end of the microphone (1) is connected to the signal acquisition end of the noise signal conditioning collector (2), and the Wi-Fi wireless data output terminal of the noise signal conditioning collector (2) is connected to the Wi-Fi connection of the computer (5). The wireless data input end is wirelessly connected, the signal output end of the fan operation parameter sensor group (3) is connected to the signal acquisition terminal group of the fan operation parameter signal acquisition conditioner (4), and the fan operation parameter signal acquisition conditioner (4) ) USB output port to connect the USB input port of the computer (5). 3. The method for testing the tone of a wind power generating set according to claim 2, characterized in that: the noise signal conditioning collector (2) is provided with a built-in piezoelectric ICP excitation constant current source (6), an AC coupler (7) , anti-aliasing filter (8), the first A/D converter (9) and Wi-Fi wireless data transmission module (10), the output end of described built-in piezoelectric ICP excitation constant current source (6) is connected described The power supply terminal of the prepolarized microphone (1), the signal output terminal of the prepolarized microphone (1) is connected to the signal acquisition terminal of the AC coupler (7), and the signal output terminal of the AC coupler (7) is connected to the The signal input end of the anti-aliasing filter (8), the signal output end of the anti-aliasing filter (8) is connected to the signal input end of the first A/D converter (9), the first A/D conversion The signal output end of the device (9) is connected to the signal input end of the Wi-Fi wireless data transmission module (10), and the signal output end of the Wi-Fi wireless data transmission module (10) is connected to the Wi-Fi of the computer (5). -Fi wireless data input end wireless connection; The fan operation parameter signal acquisition conditioner (4) is provided with a multiplexer (11), an amplification filter (12), a second A/D converter (13) and a USB data transmission module (14), the The signal output end of the fan operating parameter sensor group (3) is connected to the signal acquisition end group of the multiplexer (11), and the signal output end of the multiplexer (11) is connected to the amplifier filter (12). signal input end, the signal output end of the amplification filter (12) is connected to the signal input end of the second A/D converter (13), and the signal output end of the second A/D converter (13) is connected to the The signal input end of the USB data transmission module (14), the signal output end of the USB data transmission module (14) is connected to the USB input end of the computer (5). 4. The method for testing the tone of a wind power generating set according to claim 2, characterized in that: the noise signal conditioning collector (2) is powered by a rechargeable battery pack. 5. The method for testing the tone of a wind power generating set according to claim 2, characterized in that: the fan operating parameter sensor group (3) is composed of an electric power sensor, a wind speed sensor, a temperature sensor, an air pressure sensor and a rotational speed sensor, and the electric power The signal output end of the sensor is connected to the electric power signal acquisition end of the fan operation parameter signal acquisition conditioner (4), and the signal output end of the wind speed sensor is connected to the wind speed signal acquisition end of the fan operation parameter signal acquisition conditioner (4), The signal output end of the air temperature sensor is connected to the air temperature signal acquisition end of the fan operation parameter signal acquisition conditioner (4), and the signal output end of the air pressure sensor is connected to the air pressure signal acquisition end of the fan operation parameter signal acquisition conditioner (4). The signal output end of the speed sensor is connected to the speed signal acquisition end of the fan operation parameter signal acquisition conditioner (4). 6. The method for testing the tone of a wind power generating set according to claim 2, characterized in that: the testing software of the computer (5) is a virtual instrument.

Images