CN110135272A - It is centrifuged the quantitative identification method of pump pressure fluctuation broadband frequency - Google Patents
It is centrifuged the quantitative identification method of pump pressure fluctuation broadband frequency Download PDFInfo
- Publication number
- CN110135272A CN110135272A CN201910317017.2A CN201910317017A CN110135272A CN 110135272 A CN110135272 A CN 110135272A CN 201910317017 A CN201910317017 A CN 201910317017A CN 110135272 A CN110135272 A CN 110135272A
- Authority
- CN
- China
- Prior art keywords
- frequency
- broadband
- broadband frequency
- fitting function
- point
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/15—Correlation function computation including computation of convolution operations
- G06F17/156—Correlation function computation including computation of convolution operations using a domain transform, e.g. Fourier transform, polynomial transform, number theoretic transform
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
- G06F2218/04—Denoising
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Algebra (AREA)
- Control Of Non-Positive-Displacement Pumps (AREA)
- Complex Calculations (AREA)
Abstract
The embodiment of the present invention provides a kind of quantitative identification method for being centrifuged pump pressure fluctuation broadband frequency, comprising: by recording the transient pressure of centrifugal pump, each of obtains the frequency domain data of pressure coefficient, and extract frequency domain data broadband frequency；For each broadband frequency, it is based on sieve value iterative method, is fitted using whole maximum points of the normal distribution form to broadband frequency, obtains the centre frequency of the first fitting function；The degree of deviation is calculated according to centre frequency and peak frequency, and frequency corresponding distributional pattern type in each broadband is confirmed according to the degree of deviation；Each broadband frequency is fitted using approximating method corresponding with distributional pattern type, obtains the corresponding target fitting function of each broadband frequency, and obtains the Morphologic stereology characterization parameter of broadband frequency based on target fitting function.The embodiment of the present invention carries out quantitative identification to centrifugation pump pressure fluctuation broadband frequency by Morphologic stereology characterization parameter, and then can establish the complete Quantitative Evaluation System of centrifugal pump the fluctuating frequency of pressure characteristic.
Description
Technical field
The present invention relates to field of hydraulic machinery, more particularly, to a kind of quantitative knowledge for being centrifuged pump pressure fluctuation broadband frequency
Other method.
Background technique
In hydraulic operational process, since the stator-rotor interaction between wheel rotation, rotary part and stationary parts is made
With the influence of, factors such as cavitating flows, Karman vortex street, rotating stall, flow separation and reflux, transient state pressure inside hydraulic
Unsteady pulse characteristic is presented in power.High amplitude pressure, which is pulsed, can cause vibration and the noise of hydraulic and its system, in pressure
Even it can cause to resonate when ripple frequency is consistent with structural natural frequencies.The frequency content and width of accurate quantification identification pressure fluctuation
Value is of great significance for evaluation hydraulic operation stability.The appraisement system of pressure fluctuation frequency domain in the prior art,
It is mainly focused on and turns the single-frequencies ingredients such as frequency, leaf frequency and draft tube vortex rope.But abundant experimental results show in hydraulic
Portion, especially centrifugal pump, often there is also the cline frequency band with elevated pressures amplitude, for broadband character frequency content,
There is presently no reliable quantitative judge and evaluation methods.
Summary of the invention
To solve the above-mentioned problems, the embodiment of the present invention provides one kind and overcomes the above problem or at least be partially solved
State the quantitative identification method of the centrifugation pump pressure fluctuation broadband frequency of problem.
According to a first aspect of the embodiments of the present invention, a kind of quantitative judge side being centrifuged pump pressure fluctuation broadband frequency is provided
Method, this method comprises: obtaining the frequency domain data of pressure coefficient, and extract frequency domain data by the transient pressure of record centrifugal pump
Each of broadband frequency；For each broadband frequency, it is based on sieve value iterative method, using normal distribution form to the complete of broadband frequency
Portion's maximum point is fitted, and obtains the centre frequency of the first fitting function；Deviation is calculated according to centre frequency and peak frequency
Degree, and frequency corresponding distributional pattern type in each broadband is confirmed according to the degree of deviation；Wherein, distributional pattern type includes symmetrical state
Distributional pattern or partial velocities form, peak frequency are the corresponding frequency of amplitude of broadband frequency；Using with distributional pattern type pair
The approximating method answered is fitted each broadband frequency, obtains the corresponding target fitting function of each broadband frequency, and be based on
The Morphologic stereology characterization parameter of target fitting function acquisition broadband frequency；Wherein, the sieve value iterative method includes the following steps: to institute
It states after maximum point is fitted, if coefficient of determination R-square is less than setting accuracy threshold value, casts out whole maximum
Target Max point in point obtains remaining maximum point；The Target Max point is that pressure coefficient is less than matched curve
The point of first setting multiple；The remaining maximum point is fitted, until R-square is not less than the setting accuracy threshold
Value.
According to a second aspect of the embodiments of the present invention, a kind of electronic equipment, including memory, processor and storage are provided
On a memory and the computer program that can run on a processor, processor is realized various such as first aspect when executing program
The quantitative judge side of centrifugation pump pressure fluctuation broadband frequency provided by any possible implementation in possible implementation
Method.
The quantitative identification method of centrifugation pump pressure fluctuation broadband frequency provided in an embodiment of the present invention, determines according to the degree of deviation
The corresponding distributional pattern type of broadband frequency, so as to according to distributional pattern type by reasonable approximating method to broadband frequency into
Row fitting, obtains target fitting function, extracts the Morphologic stereology characterization parameter of broadband frequency；So as to pass through Morphologic stereology table
Sign parameter carries out quantitative identification to centrifugation pump pressure fluctuation broadband frequency, and then can establish centrifugal pump the fluctuating frequency of pressure characteristic
Complete Quantitative Evaluation System.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described.It should be evident that the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
Other attached drawings are obtained according to these figures.
Fig. 1 is the process signal of the quantitative identification method of centrifugation pump pressure fluctuation broadband frequency provided in an embodiment of the present invention
Figure；
Fig. 2 be another embodiment of the present invention provides the process of quantitative identification method of centrifugation pump pressure fluctuation broadband frequency show
It is intended to；
Fig. 3 is the time-domain diagram of nondimensionalization provided in an embodiment of the present invention；
Fig. 4 is the time-domain diagram after provided in an embodiment of the present invention plus Hanning (Chinese is peaceful) window；
Fig. 5 is pressure coefficient frequency domain figure provided in an embodiment of the present invention；
Fig. 6 is the frequency domain figure provided in an embodiment of the present invention for removing frequency multiplication；
Fig. 7 is broadband frequency positioning schematic diagram after removing white noise provided in an embodiment of the present invention；
Fig. 8 carries out first time fitting to all maximum points using normal distyribution function to be provided in an embodiment of the present invention
Schematic diagram；
Fig. 9 carries out second of fitting to remaining maximum point using normal distyribution function to be provided in an embodiment of the present invention
Schematic diagram；
Figure 10 be it is provided in an embodiment of the present invention using normal distyribution function to all maximum points in peak frequency two sides into
Went amplitude point first time fitting schematic diagram；
Figure 11 be it is provided in an embodiment of the present invention using normal distyribution function to peak frequency two sides residue maximum point into
Went the schematic diagram of second of amplitude point fitting；
Figure 12 be it is provided in an embodiment of the present invention using normal distyribution function to peak frequency two sides residue maximum point into
The schematic diagram that the third time of amplitude point of going is fitted；
Figure 13 carried out all maximum points in peak frequency two sides using exponential function to be provided in an embodiment of the present invention
The schematic diagram of the first time fitting of amplitude point；
Figure 14 carried out peak frequency two sides residue maximum point using exponential function to be provided in an embodiment of the present invention
The schematic diagram of second of fitting of amplitude point；
Figure 15 carried out peak frequency two sides residue maximum point using exponential function to be provided in an embodiment of the present invention
The schematic diagram of the third time fitting of amplitude point；
Figure 16 is frequency quantitative judge schematic diagram in broadband provided in an embodiment of the present invention；
Figure 17 is the entity structure schematic diagram of electronic equipment provided in an embodiment of the present invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical solution in the embodiment of the present invention is explicitly described, it is clear that described embodiment is the present invention
A part of the embodiment, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art are not having
Every other embodiment obtained under the premise of creative work is made, shall fall within the protection scope of the present invention.
The embodiment of the present invention provides a kind of quantitative identification method for being centrifuged pump pressure fluctuation broadband frequency, referring to Fig. 1, this method
Including but not limited to following steps:
Step 100, the transient pressure by recording centrifugal pump, obtain the frequency domain data of pressure coefficient, and extract frequency domain number
According to each of broadband frequency.
Wherein it is possible to using the transient pressure of the method for experiment or numerical simulation record centrifugal pump, it can be according to practical need
It asks and reasonably selects sample frequency and record time.For example, sample frequency can be 1024Hz, the record time can be 10s, that
, the data point interception number of record is 2N, N is positive integer；It such as can be 8192.According to transient pressure, can be pressed
The frequency domain data of force coefficient is specifically as follows pressure coefficient frequency domain figure.It, can be fixed by setting a threshold value based on frequency domain data
Each of position and extraction frequency domain data broadband frequency.
Step 200, for each broadband frequency, be based on sieve value iterative method, using normal distribution form to the complete of broadband frequency
Portion's maximum point is fitted, and obtains the centre frequency of the first fitting function.
Specifically, after being extracted each broadband frequency in step 100, for each broadband frequency, located as follows
Reason: all maximum points of broadband frequency are obtained, are fitted using normal distribution form, normal distribution morphic function can indicate
Are as follows:
Wherein, fiIt is relative frequency, A0For fitting function amplitude, a fitting function centre frequency, σ is fitting function standard
Difference.
The first fitting function can be obtained after fitting, and can further obtain the centre frequency of the first fitting function.
Step 300, the degree of deviation is calculated according to centre frequency and peak frequency, and each broadband frequency is confirmed according to the degree of deviation
Corresponding distributional pattern type；Wherein, distributional pattern type includes symmetrical state distributional pattern or partial velocities form, peak frequency
For the corresponding frequency of highest amplitude in the frequency of broadband.
Specifically, can define the corresponding frequency of highest amplitude in the frequency of broadband is peak frequency.Is obtained in step 200
After the centre frequency of one fitting function, the degree of deviation between peak frequency and centre frequency can be calculated, the degree of deviation reflects the two
Between extent of deviation.And the distributional pattern type of each broadband frequency can be further determined according to the degree of deviation of each broadband frequency.
Distributional pattern type can be one in symmetrical state distributional pattern and partial velocities form.
Step 400 is fitted each broadband frequency using approximating method corresponding with distributional pattern type, obtains every
The corresponding target fitting function of one broadband frequency, and the Morphologic stereology characterization parameter based on target fitting function acquisition broadband frequency.
It specifically, can be to distributional pattern kind after obtaining the corresponding distributional pattern type of each broadband frequency in step 300
Class is the broadband frequency of symmetrical state distributional pattern using a kind of approximating method, and to distributional pattern type is the width of partial velocities form
Band frequency uses another kind of approximating method, may finally obtain the corresponding target fitting function of each broadband frequency.Target is fitted letter
Number should be the fitting function for capableing of the maximum probability of maximum envelope broadband frequency.It is quasi- so as to be based further on target
Function is closed to extract the Morphologic stereology characterization parameter of broadband frequency.Concrete type of the embodiment of the present invention to Morphologic stereology characterization parameter
It is not construed as limiting.
Wherein, the sieve value iterative method in above-mentioned steps 200 includes but is not limited to following steps:
Step 1 after being fitted to maximum point, if coefficient of determination R-square is less than setting accuracy threshold value, is cast out
Target Max point in whole maximum points obtains remaining maximum point；Target Max point is that pressure coefficient is less than fitting
The point of first setting multiple of curve.
Step 2 is fitted remaining maximum point, until R-square is not less than setting accuracy threshold value.
Specifically, fitting precision is measured using coefficient of determination R-square.If R-square is less than setting accuracy threshold value
(such as 0.9), then cast out Target Max point, and Target Max point is pressure coefficient less than the first setting multiple (such as 0.9
The point of matched curve again).And to remaining maximum point carry out respective type distributional pattern (including normal distribution form be fitted
Or exponential distribution form) fitting, until fitting precision be not less than setting accuracy threshold value.After the completion of fitting, it can get finally
Remaining maximum point corresponding to fitting function and final fitting function.
Therefore, above-mentioned sieve value iterative method can be used, normal distribution form is carried out to whole maximum points in step 200
Fitting obtains the first fitting function and the corresponding remaining maximum point of the first fitting function.
The quantitative identification method of centrifugation pump pressure fluctuation broadband frequency provided in an embodiment of the present invention, determines according to the degree of deviation
The corresponding distributional pattern type of broadband frequency, so as to according to distributional pattern type by reasonable approximating method to broadband frequency into
Row fitting, obtains target fitting function, extracts the Morphologic stereology characterization parameter of broadband frequency；So as to pass through Morphologic stereology table
Sign parameter carries out quantitative identification to centrifugation pump pressure fluctuation broadband frequency, and then can establish centrifugal pump the fluctuating frequency of pressure characteristic
Complete Quantitative Evaluation System.
Content based on the above embodiment provides a kind of transient state by recording centrifugal pump as a kind of alternative embodiment
Pressure, the method for obtaining the frequency domain data of pressure coefficient, including but not limited to following steps:
Step 101, the transient pressure for recording centrifugal pump carry out nondimensionalization to transient pressure, obtain pressure coefficient.
Wherein, nondimensionalization, pressure coefficient representation method are carried out to transient pressure are as follows:
Wherein, CpIt is pressure coefficient, piIt is the pressure of transient state, [Pa],It is average pressure, [Pa], u exit edge of blade
Circumferential speed, [m/s].
Step 102 carries out frequency-domain transform after adding Hanning window function to pressure coefficient in time-domain, and on frequency domain
Pressure coefficient is modified, and is carried out nondimensionalization to frequency, is obtained the frequency domain data of pressure coefficient.
Specifically, Hanning window function is added to pressure coefficient in entire time-domain, is pressed using Fast Fourier Transform (FFT)
Force coefficient frequency domain figure (i.e. the frequency domain data of pressure coefficient)；By the pressure coefficient on frequency domain multiplied by corrected parameter (such as)
It is modified.Nondimensionalization, relative frequency representation method are carried out to frequency are as follows:
fc=fp/fr
Wherein, fcIt is relative frequency, fpIt is the actual frequency of pressure fluctuation, [Hz], frIt is impeller speed, [Hz].
Content based on the above embodiment obtains pressure coefficient as a kind of alternative embodiment in above-mentioned steps 102
Further include following steps after frequency domain data:
The integer frequency for turning frequency and turning frequency in step 103, removal frequency domain data.In other words, it finds and turns in frequency domain figure
The multiple of frequency, and removed.
Step 104, the average value for obtaining pressure coefficient in frequency domain data, and using average value as threshold value, remove frequency domain number
Less than the point of the pressure coefficient of threshold value in, and the broadband frequency is removed by adjusting the size of threshold value in removal process
The white noise of two sides.
Specifically, the average value P of pressure coefficient in frequency domain data is obtainedave, enable PaveAs threshold value, removes in frequency domain figure and press
Force coefficient is less than the point of threshold value.In addition, threshold value increases to 1.1 times, removes again if broadband frequency two sides are there are also white noise presence
Pressure coefficient is less than the point of threshold value in frequency domain data, until not having white noise presence.
Content based on the above embodiment provides in a kind of step 200 using normal distribution as a kind of alternative embodiment
The method that form is fitted whole maximum points of broadband frequency, including but not limited to following steps: the complete of broadband frequency is obtained
Portion's maximum point, and it is based on sieve value iterative method, whole maximum points are fitted using normal distribution form, it is quasi- to obtain first
Close function.
Content based on the above embodiment provides a kind of according to centre frequency and peak frequency as a kind of alternative embodiment
Rate calculates the degree of deviation, and the method for the corresponding distributional pattern type of each broadband frequency is confirmed according to the degree of deviation, including but unlimited
In following steps:
Difference between step 301, calculating centre frequency and peak frequency, according to difference and the standard of the first fitting function
Difference calculates the degree of deviation.
Wherein, the degree of deviation can be the ratio between the absolute value of above-mentioned difference and the standard deviation of certain multiple (such as 6 times)
Value.
If step 302, the degree of deviation are less than degree of deviation threshold value, confirm the distributional pattern type of broadband frequency for symmetrical state distribution
Form；Otherwise, the distributional pattern type of broadband frequency is partial velocities form.
Wherein, degree of deviation threshold value can be 3%.That is, distributional pattern type is symmetrical state point if the degree of deviation is less than 3%
Otherwise cloth form is partial velocities form.
Content based on the above embodiment, as a kind of alternative embodiment, distributional pattern type is symmetrical state distributional pattern,
Correspondingly, it provides a kind of use approximating method corresponding with distributional pattern type to be fitted each broadband frequency, obtain every
The method of the corresponding target fitting function of one broadband frequency, including but not limited to following steps:
Step 401, the amplitude point for obtaining the first fitting function, and it is by boundary line of centre frequency that the whole of broadband frequency is very big
Value point is divided into left and right two parts；Based on sieve value iterative method, using exponential distribution form respectively to the two-part maximum point in left and right
The fitting of the amplitude point of the first fitting function was carried out, the second fitting function is obtained.
Specifically, it is symmetrical state distributional pattern if broadband frequency, obtains the amplitude point of the first fitting function.With centre frequency
All maximum points of broadband frequency are divided into left and right two parts for boundary line, using exponential distribution form respectively to the pole of the right and left
Big value point carried out the fitting of amplitude point, and specifically sieve value iterative method can be used in fitting, and exponential distribution morphic function may be expressed as:
Wherein, b and c is the coefficient of exponential fitting, f0For peak frequency.
It can get the second fitting function and the corresponding remaining maximum point of the second fitting function after the completion of fitting.
If it is corresponding that the quantity of the corresponding remaining maximum point of step 402, the first fitting function is greater than the second fitting function
The quantity of remaining maximum point, then using the first fitting function as the target fitting function of broadband frequency；Otherwise, letter is fitted by second
Target fitting function of the number as broadband frequency.
It is greater than the corresponding remaining pole of the second fitting function in the quantity of the corresponding remaining maximum point of the first fitting function
When the quantity of big value point, using normal distribution form fitting broadband frequency；Otherwise, using exponential distribution form fitting broadband frequency.
Content based on the above embodiment, as a kind of alternative embodiment, distributional pattern type is partial velocities form；Phase
Ying Di provides a kind of use approximating method corresponding with distributional pattern type and is fitted to each broadband frequency, obtains each
The method of the corresponding target fitting function of a broadband frequency, including but not limited to following steps:
Step 401 ', obtain the amplitude point of broadband frequency, and be boundary line by whole maximum points of broadband frequency using peak frequency
It is divided into left and right two parts.
Specifically, if broadband frequency is partial velocities form, the amplitude point of broadband frequency is taken, and will by boundary line of peak frequency
All maximum points of broadband frequency are divided into left and right two parts.
Step 402 ', be based on sieve value iterative method, using normal distribution form respectively to the two-part maximum point in left and right into
Went broadband frequency amplitude point fitting, obtain third fitting function；And it is based on sieve value iterative method, using exponential distribution shape
State carried out the fitting of the amplitude point of broadband frequency to the two-part maximum point in left and right respectively, obtained the 4th fitting function.
Specifically, the quasi- of amplitude point was carried out to the maximum point of the right and left respectively using normal distribution form first
It closes, fitting uses sieve value iterative method, obtains third fitting function.Then using exponential distribution form respectively to the pole of the right and left
Big value point carried out the fitting of amplitude point, and fitting uses sieve value iterative method, obtained the 4th fitting function.It should be noted that this
Inventive embodiments do not make the sequencing for being fitted and being fitted using exponential distribution form using normal distribution form
It limits.
Step 403 ' if, that the quantity of the corresponding remaining maximum point of third fitting function is greater than the 4th fitting function is corresponding
The quantity of remaining maximum point, then using third fitting function as the target fitting function of broadband frequency；Otherwise, letter is fitted by the 4th
Target fitting function of the number as broadband frequency.
Specifically, since fitting is using sieve value iterative method, the fitting function after the completion of fitting is corresponding with difference
The maximum point of quantity.Quantity by comparing the corresponding remaining maximum point of third fitting function is corresponding with the 4th fitting function
Remaining maximum point quantity, it is after obtaining relatively as a result, to use corresponding distribution shape according to the result after comparison
State is fitted broadband frequency.It is corresponding surplus to be greater than the 4th fitting function in the quantity of the corresponding remaining maximum point of third fitting function
When the quantity of remaining maximum point, using normal distribution form fitting broadband frequency；Otherwise, broadband is fitted using exponential distribution form
Frequently.
Content based on the above embodiment, as a kind of alternative embodiment, Morphologic stereology characterization parameter includes: bandwidth, band
At least one of wide coefficient, bandwidth center frequency and broadband frequency amplitude；Correspondingly, it provides in a kind of step 400 based on target
The method that fitting function obtains the Morphologic stereology characterization parameter of broadband frequency, including but not limited to following steps:
The straight line for making the amplitude for the broadband frequency that pressure coefficient is the second setting multiple, obtains straight line and target fitting function
The distance between two intersection points and two intersection points；Bandwidth is the distance between two intersection points；Bandwidth factor is bandwidth and broadband frequency
Amplitude between ratio；If distributional pattern type is symmetrical state distribution, bandwidth center frequency is in target fitting function
Frequency of heart；If distributional pattern type is partial velocities, bandwidth center frequency is the peak frequency of broadband frequency；The broadband frequency width
Value is the maximum value of the target fitting function.
Wherein, the second setting multiple can be 0.5 times, and the embodiment of the present invention is not construed as limiting this.It is equal to as pressure coefficient
The straight line of 0.5 times of amplitude obtains straight line at a distance from two intersection points and point-to-point transmission of fitting function, is broadband by the distance definition
The ratio of the distance and amplitude is defined as bandwidth factor by the bandwidth of frequency, by the centre frequency or partial velocities of symmetrical state distribution
Peak frequency be defined as broadband centre frequency；By the maximum value of target fitting function broadband frequency amplitude the most.
In order to be illustrated to above-described embodiment, another embodiment of the present invention provides a kind of centrifugation pump pressure fluctuation broadband frequencies
Quantitative identification method, referring to fig. 2, this method includes but is not limited to following steps:
S1, the transient pressure for recording water pump, sample frequency are not less than 1024Hz, and the record time is 10s, the data point of interception
Number is 2N, N=13, to transient pressure carry out nondimensionalization, obtain pressure coefficient；
S2, Hanning (Chinese is peaceful) window function is added to pressure coefficient in entire time-domain, and carries out frequency-domain transform, and right
The pressure coefficient of frequency domain is modified, and is carried out nondimensionalization to frequency, is obtained relative frequency；
S3, all turns of frequencies and its integer frequency in frequency domain are found, and removed it；
S4, the average value P for obtaining pressure coefficient in frequency domainave, enable PaveAs threshold value, it is small to remove pressure coefficient in frequency domain figure
In the point of threshold value, if broadband frequency two sides, there are also white noise presence, threshold value increases to 1.1 times, until not having white noise presence；
S5, threshold value is adjusted using method in S4, positions all broadband frequencies, and each broadband frequency is individually extracted；
S6, all maximum points of broadband frequency are obtained, is fitted using normal distyribution function, using coefficient of determination R-
Square measures fitting precision, if R-square less than 0.9, casts out point of the pressure coefficient less than 0.9 times of matched curve, and right
Remaining maximum point carries out normal distyribution function fitting, until fitting precision is not less than 0.9 (hereinafter referred to as sieve value iterative method)；
S7, the centre frequency a=11.8 for obtaining fitting function and standard deviation sigma=0.1984 define broadband frequency highest amplitude
Corresponding frequency is peak frequency f0=11.75, the difference of peak frequency and centre frequency is calculated, the definition degree of deviation is above-mentioned difference
The ratio of the absolute value of value and 6 times of standard deviations, if the degree of deviation is otherwise partial velocities for symmetrical state distribution less than 3%；
S8, if symmetrical state distributional pattern, obtain the amplitude point of fitting function, and be boundary line by broadband frequency using centre frequency
All maximum points be divided into left and right two parts, width was carried out to the maximum point of the right and left respectively using exponential distribution form
It is worth the fitting of point, sieve value iterative method is used when fitting；
If the number of remaining maximum point is fitted broadband frequency using normal distribution form, otherwise more than in S8 in S9, S6
Using exponential distribution form；
S10, if partial velocities form, take the amplitude point (11.75,0.0017) of broadband frequency, and using peak frequency as boundary
All maximum points of broadband frequency are divided into left and right two parts by line, using normal distribution form respectively to the maximum of the right and left
Point carried out the fitting of amplitude point, and when fitting uses sieve value iterative method；
S11, if partial velocities form, take the amplitude point (11.75,0.0017) of broadband frequency, and using peak frequency as boundary
All maximum points of broadband frequency are divided into left and right two parts by line, using exponential distribution form respectively to the maximum of the right and left
Point carried out the fitting of amplitude point, and when fitting uses sieve value iterative method；
If the number of remaining maximum point is fitted broadband frequency using normal distribution form more than in S11 in S12, S10,
Otherwise exponential distribution form is used；
If S13, broadband frequency are symmetrical state distributional pattern, then used using approximating method in S9 if partial velocities form
Approximating method in S12 makees the straight line that pressure coefficient is equal to 0.5 times of amplitude, obtains two intersection points and two of straight line and fitting function
The distance definition is the bandwidth of broadband frequency, the ratio of the distance and amplitude is defined as bandwidth factor by the distance between point, will be right
The peak frequency of the centre frequency and partial velocities that claim state distribution is defined as broadband centre frequency, and the maximum value of fitting function is determined
Justice is broadband frequency amplitude.
One presented below combines the example of the indoor pressure fluctuation data of double suction centrifugal pump pressure to the embodiment of the present invention
The method of offer is further described:
As shown in figure 3, sample frequency is 1024Hz, the data point interception number of record is 8192, is carried out to pressure immeasurable
Guiding principle obtains pressure fluctuation time-domain diagram.
As shown in figure 4, adding Hanning (Chinese is peaceful) window function in entire time-domain to prevent energy leakage.
As shown in figure 5, obtain pressure coefficient frequency domain figure by Fast Fourier Transform (FFT), by the pressure coefficient of frequency domain multiplied byIt is modified, nondimensionalization is carried out to frequency, the frequency multiplication of all amplitude is found, determines the amplitude of white noise.
As shown in fig. 6, removal turns frequency and its integer frequency.
As shown in fig. 7, obtaining the average value P of pressure coefficient in frequency domainave, enable PaveAs threshold value, removes in frequency domain figure and press
Force coefficient is less than the point of threshold value, and the relative frequency ranges for obtaining first broadband frequency are 11.0-12.6.
As shown in figure 8, finding all maximum points of broadband frequency, and carries out normal distyribution function and be fitted for the first time, R-
Square=0.6987；As shown in figure 9, removing all maximum points less than 0.9 times of matched curve, and carry out normal distribution letter
Second of fitting of number, R-square=0.9155；Normal distribution centre frequency a=11.8, broadband frequency highest amplitude point are corresponding
Frequency f0=11.75, standard deviation sigma=0.1984, the degree of deviation 4.2% determines that broadband frequency is partial velocities form.
As shown in Figure 10, it takes the amplitude point (11.75,0.0017) of broadband frequency, and is boundary line by broadband frequency using peak frequency
All maximum points be divided into left and right two parts, amplitude was carried out to the right and left maximum point respectively using normal distribution form
The Function Fitting of point, left side R-square=0.1611, the right R-square=0.4655；As shown in figure 11, remove all small
In the maximum point of 0.9 times of matched curve, and carries out second of normal distyribution function and be fitted, left side R-square=0.7962,
The right R-square=0.8752；As shown in figure 12, remove all maximum points less than 0.9 times of matched curve, and carry out just
State distribution function third time is fitted, left side R-square=0.9211, the right R-square=0.9434.
As shown in figure 13, it takes the amplitude point (11.75,0.0017) of broadband frequency, and is boundary line by broadband frequency using peak frequency
All maximum points be divided into left and right two parts, amplitude was carried out to the right and left maximum point respectively using exponential distribution form
The Function Fitting of point, left and right R-square=0.5552, the right R-square=0.7457；As shown in figure 14, remove all small
In the maximum point of 0.9 times of matched curve, goes forward side by side second of row index function and be fitted, left and right R-square=0.0.8936 is right
Side R-square=0.9254；As shown in figure 15, remove all maximum points less than 0.9 times of matched curve, row index of going forward side by side
Function third time is fitted, left side R-square=0.9487, the right R-square=0.9481.
It is 58 that normal distyribution function, which is fitted remaining maximum point number, and exponential function is fitted remaining maximum point number and is
26, final choice normal distyribution function as shown in figure 12 is fitted broadband frequency.
As shown in figure 16, it is straight line Cp=0.5A0, straight line is obtained at a distance from the intersection point and point-to-point transmission of fitting function, by this
Distance definition is the bandwidth W=0.4115 of broadband frequency, and the ratio of the distance and amplitude is defined as bandwidth factor Wc=242.06,
The peak frequency of partial velocities is defined as broadband centre frequency fmid=11.75, broadband frequency amplitude A0=0.0017.
To sum up, the quantitative identification method of centrifugation pump pressure fluctuation broadband frequency provided in an embodiment of the present invention, including, record pressure
Power ripple factor time-domain signal；Obtain dimensionless frequency-region signal；Eliminate fundamental frequency and its frequency multiplication；Threshold value eliminates white noise, and
Successively position broadband frequency range；It is calculated based on sieve value iterative method using the maximum point of normal distribution morphic function fitting broadband frequency
Otherwise broadband is frequently partial velocities shape if being judged to meeting symmetrical state distributional pattern less than 3% with the degree of deviation of fitting function
State；For the broadband frequency of symmetrical state distributional pattern, it is based on sieve value iterative method, the maximum point on centre frequency both sides is used respectively
The fitting of exponential distribution form, compares the quantity of normal distribution fitting and exponential fitting residue maximum point, and selects remaining very big
The more approximating method of value point；For the broadband frequency of partial velocities form, it is based on sieve value iterative method, respectively to peak frequency both sides
Maximum point is fitted using the fitting of normal distribution form with exponential distribution form, and the approximating method more than remaining maximum point is selected；
Determine bandwidth, bandwidth factor, broadband centre frequency and broadband frequency amplitude.This method determines centrifugation pump pressure fluctuation broadband frequency
The identification of amount, and then the complete Quantitative Evaluation System of centrifugal pump the fluctuating frequency of pressure characteristic can be established.
The embodiment of the invention provides a kind of electronic equipment, and as shown in figure 17, which includes: processor
(processor) 501, communication interface (Communications Interface) 502, memory (memory) 503 and communication
Bus 504, wherein processor 501, communication interface 502, memory 503 complete mutual communication by communication bus 504.
Processor 501 can call the computer program that can be run on memory 503 and on processor 501, to execute above-mentioned each reality
Apply the quantitative identification method of the centrifugation pump pressure fluctuation broadband frequency of example offer, for example, by the transient state pressure for recording centrifugal pump
Power each of obtains the frequency domain data of pressure coefficient, and extracts frequency domain data broadband frequency；For each broadband frequency, base
It in sieve value iterative method, is fitted using whole maximum points of the normal distribution form to broadband frequency, obtains the first fitting function
Centre frequency；The degree of deviation is calculated according to centre frequency and peak frequency, and confirms that each broadband frequency is corresponding according to the degree of deviation
Distributional pattern type；Wherein, distributional pattern type includes symmetrical state distributional pattern or partial velocities form, and peak frequency is width
The corresponding frequency of amplitude with frequency；Each broadband frequency is fitted using approximating method corresponding with distributional pattern type,
The corresponding target fitting function of each broadband frequency is obtained, and obtains the quantitative Morphologic stereology of broadband frequency based on target fitting function
Characterization parameter.
In addition, the logical order in above-mentioned memory 503 can be realized by way of SFU software functional unit and conduct
Independent product when selling or using, can store in a computer readable storage medium.Based on this understanding, originally
Substantially the part of the part that contributes to existing technology or the technical solution can be in other words for the technical solution of invention
The form of software product embodies, which is stored in a storage medium, including some instructions to
So that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation of the present invention
The all or part of the steps of example method.And storage medium above-mentioned includes: USB flash disk, mobile hard disk, read-only memory (ROM, Read-
Only Memory), random access memory (RAM, Random Access Memory), magnetic or disk etc. are various can be with
Store the medium of program code.
The embodiment of the present invention also provides a kind of non-transient computer readable storage medium, is stored thereon with computer program,
The computer program is implemented to carry out the various embodiments described above and provides determining for centrifugation pump pressure fluctuation broadband frequency when being executed by processor
Measure recognition methods, for example, by recording the transient pressure of centrifugal pump, obtain the frequency domain data of pressure coefficient, and extract frequency
Each of numeric field data broadband frequency；For each broadband frequency, it is based on sieve value iterative method, using normal distribution form to broadband
Whole maximum points of frequency are fitted, and obtain the centre frequency of the first fitting function；According to centre frequency and peak frequency meter
The degree of deviation is calculated, and frequency corresponding distributional pattern type in each broadband is confirmed according to the degree of deviation；Wherein, distributional pattern type includes
Symmetrical state distributional pattern or partial velocities form, peak frequency are the corresponding frequency of amplitude of broadband frequency；Using with distributional pattern
The corresponding approximating method of type is fitted each broadband frequency, obtains the corresponding target fitting function of each broadband frequency,
And the quantitative Morphologic stereology characterization parameter of broadband frequency is obtained based on target fitting function.
The embodiments such as electronic equipment described above are only schematical, wherein unit as illustrated by the separation member
It may or may not be physically separated, component shown as a unit may or may not be physics list
Member, it can it is in one place, or may be distributed over multiple network units.It can be selected according to the actual needs
In some or all of the modules achieve the purpose of the solution of this embodiment.Those of ordinary skill in the art are not paying creativeness
Labour in the case where, it can understand and implement.
Through the above description of the embodiments, those skilled in the art can be understood that each embodiment can
It realizes by means of software and necessary general hardware platform, naturally it is also possible to pass through hardware.Based on this understanding, on
Stating technical solution, substantially the part that contributes to existing technology can be embodied in the form of software products in other words, should
Computer software product may be stored in a computer readable storage medium, such as ROM/RAM, magnetic disk, CD, including several fingers
It enables and using so that a computer equipment (can be personal computer, server or the network equipment etc.) executes each implementation
Certain Part Methods of example or embodiment.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；
And these are modified or replaceed, technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (9)
1. a kind of quantitative identification method for being centrifuged pump pressure fluctuation broadband frequency characterized by comprising
By recording the transient pressure of centrifugal pump, the frequency domain data of pressure coefficient is obtained, and extract every in the frequency domain data
One broadband frequency；
For the frequency of broadband described in each, it is based on sieve value iterative method, using normal distribution form to whole poles of the broadband frequency
Big value point is fitted, and obtains the centre frequency of the first fitting function；
The degree of deviation is calculated according to the centre frequency and peak frequency, and each described broadband frequency is confirmed according to the degree of deviation
Corresponding distributional pattern type；Wherein, the distributional pattern type includes symmetrical state distributional pattern or partial velocities form, described
Peak frequency is the corresponding frequency of highest amplitude in the broadband frequency；
Each described broadband frequency is fitted using approximating method corresponding with the distributional pattern type, obtains each
The corresponding target fitting function of broadband frequency, and obtain based on the target fitting function Morphologic stereology table of the broadband frequency
Levy parameter；
Wherein, the sieve value iterative method includes the following steps:
After being fitted to the maximum point, if coefficient of determination R-square is less than setting accuracy threshold value, cast out described complete
Target Max point in portion's maximum point obtains remaining maximum point；The Target Max point is pressure coefficient less than quasi-
Close the point of the first setting multiple of curve；
The remaining maximum point is fitted, until R-square is not less than the setting accuracy threshold value.
2. the quantitative identification method of centrifugation pump pressure fluctuation broadband frequency according to claim 1, which is characterized in that described logical
The transient pressure of overwriting centrifugal pump obtains the frequency domain data of pressure coefficient, comprising:
The transient pressure for recording the centrifugal pump carries out nondimensionalization to the transient pressure, obtains the pressure coefficient；
Frequency-domain transform is carried out after adding Hanning window function to the pressure coefficient in time-domain, and to the pressure on frequency domain
Coefficient is modified, and is carried out nondimensionalization to frequency, is obtained the frequency domain data of the pressure coefficient.
3. the quantitative identification method of centrifugation pump pressure fluctuation broadband frequency according to claim 2, which is characterized in that obtain institute
After the frequency domain data for stating pressure coefficient, further includes:
Remove the integer frequency for turning frequency and described turn of frequency in the frequency domain data；
The average value of pressure coefficient described in the frequency domain data is obtained, and using the average value as threshold value, removes the frequency
Less than the point of the pressure coefficient of the threshold value in numeric field data, and by adjusting the size of the threshold value, removal in removal process
The white noise of the two sides of the broadband frequency.
4. the quantitative identification method of centrifugation pump pressure fluctuation broadband frequency according to claim 1, which is characterized in that according to institute
It states centre frequency and peak frequency calculates the degree of deviation, and the corresponding distribution frequently of each described broadband is confirmed according to the degree of deviation
Form type, comprising:
The difference between the centre frequency and the peak frequency is calculated, according to the difference and first fitting function
Standard deviation calculates the degree of deviation；
If the degree of deviation is less than degree of deviation threshold value, confirm that the distributional pattern type of the broadband frequency is that symmetrical state is distributed shape
State；Otherwise, the distributional pattern type of the broadband frequency is partial velocities form.
5. the quantitative identification method of centrifugation pump pressure fluctuation broadband frequency according to claim 1, which is characterized in that described point
Cloth form type is symmetrical state distributional pattern；
Correspondingly, each described broadband frequency is fitted using approximating method corresponding with the distributional pattern type, is obtained
Obtain each described broadband corresponding target fitting function frequently, comprising:
The amplitude point of first fitting function is obtained, and by boundary line of the centre frequency that the whole of the broadband frequency is very big
Value point is divided into left and right two parts；It is two-part to left and right described very big respectively using exponential distribution form based on sieve value iterative method
Value point carried out the fitting of the amplitude point of first fitting function, obtained the second fitting function；
If the quantity of the corresponding remaining maximum point of first fitting function is greater than the corresponding residue of second fitting function
The quantity of maximum point, then using first fitting function as the target fitting function of the broadband frequency；Otherwise, by described
Target fitting function of two fitting functions as the broadband frequency.
6. the quantitative identification method of centrifugation pump pressure fluctuation broadband frequency according to claim 1, which is characterized in that described point
Cloth form type is partial velocities form；
Correspondingly, each described broadband frequency is fitted using approximating method corresponding with the distributional pattern type, is obtained
Obtain each described broadband corresponding target fitting function frequently, comprising:
The amplitude point of the broadband frequency is obtained, and divides whole maximum points of the broadband frequency using the peak frequency as boundary line
For left and right two parts；
Based on sieve value iterative method, using normal distribution form, the maximum point two-part to left and right carried out the width respectively
The fitting of amplitude point with frequency obtains third fitting function；And it is based on sieve value iterative method, distinguished using exponential distribution form
The maximum point two-part to left and right carried out the fitting of the amplitude point of the broadband frequency, obtained the 4th fitting function；
If the quantity of the corresponding remaining maximum point of the third fitting function is greater than the corresponding residue of the 4th fitting function
The quantity of maximum point, then using the third fitting function as the target fitting function of the broadband frequency；Otherwise, by described
Target fitting function of four fitting functions as the broadband frequency.
7. the quantitative identification method of centrifugation pump pressure fluctuation broadband frequency according to claim 1, which is characterized in that the shape
State quantitatively characterizing parameter includes: at least one of bandwidth, bandwidth factor, broadband centre frequency and broadband frequency amplitude；
Correspondingly, the Morphologic stereology characterization parameter that the broadband frequency is obtained based on the target fitting function, comprising:
The straight line for making the amplitude for the broadband frequency that pressure coefficient is the second setting multiple, obtains the straight line and the target is fitted letter
The distance between two several intersection points and described two intersection points；The bandwidth is the distance between described two intersection points；The band
Ratio of the wide coefficient between the bandwidth and the broadband frequency amplitude；If the distributional pattern type is symmetrical state distribution,
The bandwidth center frequency is the centre frequency of the target fitting function；If the distributional pattern type is partial velocities,
The bandwidth center frequency is the peak frequency of the broadband frequency；The broadband frequency amplitude is the maximum of the target fitting function
Value.
8. a kind of electronic equipment including memory, processor and stores the calculating that can be run on a memory and on a processor
Machine program, which is characterized in that the processor realizes the centrifugal pump as described in any one of claim 1 to 7 when executing described program
The step of quantitative identification method of pressure fluctuation broadband frequency.
9. a kind of non-transient computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer journey
The quantitative judge side that pump pressure fluctuation broadband frequency is centrifuged as described in any one of claim 1 to 7 is realized when sequence is executed by processor
The step of method.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910317017.2A CN110135272B (en) | 2019-04-19 | 2019-04-19 | Quantitative identification method for pressure pulsation broadband frequency of centrifugal pump |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910317017.2A CN110135272B (en) | 2019-04-19 | 2019-04-19 | Quantitative identification method for pressure pulsation broadband frequency of centrifugal pump |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110135272A true CN110135272A (en) | 2019-08-16 |
| CN110135272B CN110135272B (en) | 2020-12-25 |
Family
ID=67570412
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910317017.2A Active CN110135272B (en) | 2019-04-19 | 2019-04-19 | Quantitative identification method for pressure pulsation broadband frequency of centrifugal pump |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110135272B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105134618A (en) * | 2015-09-10 | 2015-12-09 | 中国农业大学 | Centrifugal pump regulating method based on pressure pulsation time-frequency characteristic analysis |
| CN105889136A (en) * | 2016-04-22 | 2016-08-24 | 江苏大学 | Centrifugal pump improvement method used for reducing pressure pulse in centrifugal pump |
-
2019
- 2019-04-19 CN CN201910317017.2A patent/CN110135272B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105134618A (en) * | 2015-09-10 | 2015-12-09 | 中国农业大学 | Centrifugal pump regulating method based on pressure pulsation time-frequency characteristic analysis |
| CN105889136A (en) * | 2016-04-22 | 2016-08-24 | 江苏大学 | Centrifugal pump improvement method used for reducing pressure pulse in centrifugal pump |
Non-Patent Citations (1)
| Title |
|---|
| 郭晓乐等: "一种基于简正波模态消频散变换的声源距离深度估计方法", 《物理学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110135272B (en) | 2020-12-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107424441B (en) | Based on Hotelling's T2Aircraft track inflection point detection and estimation method of statistics | |
| CN109840495A (en) | A kind of axis frequency line enhancement method under Low SNR | |
| CN103955922A (en) | Method for detecting flaws of printed fabric based on Gabor filter | |
| CN104268883A (en) | Time-frequency spectrum curve extracting method based on edge detection | |
| CN111062425A (en) | Unbalanced data set processing method based on C-K-SMOTE algorithm | |
| CN108269274A (en) | Method for registering images based on Fourier transformation and Hough transform | |
| CN110232371A (en) | High-precision HRRP Radar Multi Target recognition methods based on small sample | |
| CN111046735A (en) | Lane line point cloud extraction method, electronic device and storage medium | |
| CN106409311A (en) | Refrain extracting apparatus and method | |
| CN110135272A (en) | It is centrifuged the quantitative identification method of pump pressure fluctuation broadband frequency | |
| CN111505598A (en) | Three-feature joint detection device and method based on FRFT domain | |
| CN114896554A (en) | Frequency modulation signal frequency range and bandwidth estimation method based on spectral feature extraction | |
| CN107884752A (en) | It is a kind of based on the external illuminators-based radar of compressed sensing to object detection method | |
| CN111901268A (en) | BPSK/QPSK signal modulation identification method based on frequency spectrum rearrangement and Gumbel distribution fitting test | |
| CN110275150A (en) | The variable accelerated motion target correlative accumulation method being fitted based on empirical mode decomposition and iteration endpoint | |
| CN115564324A (en) | Electric power fingerprint identification method, device and equipment based on event detection | |
| CN109323757B (en) | A method of estimation bubble population is to propeller sound source characteristics frequency inhibiting effect | |
| CN104199742A (en) | Method for accurately dividing blade cross section character point cloud | |
| CN113219333A (en) | Frequency spectrum parameter processing method during motor fault diagnosis | |
| CN110826904B (en) | Data processing method and device for fan, processing equipment and readable storage medium | |
| CN104569770B (en) | A kind of partial discharge time-frequency composite character parameter extracting method for extra-high-tension cable | |
| CN109507645B (en) | Method and device for extracting pulse description words | |
| CN112417774A (en) | Method and device for identifying rotating stall of compressor and computer equipment | |
| CN116962128B (en) | Method, device, equipment and storage medium for analyzing frequency shift signal | |
| CN116384009B (en) | Simulation prediction method, device and equipment for energy efficiency level of canned motor pump |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |