CN102721924B - Fault early warning method of wind generating set - Google Patents

Abstract

The present invention provides a kind of fault early warning method of wind power generating set, and it comprises the following steps: 10) collect the actual temperature value R of the monitored point N times and the parameter data related to temperature in the sampling cycle; 20) collect the The actual temperature value R of the monitored point and the parameter data related to the temperature are input into a predetermined temperature model, and the predicted temperature value of the monitored point is calculated, and the predicted temperature value is used as the theoretical temperature value P of the monitored point; 30) Obtain the temperature deviation DEV, temperature mean square error MSE and correlation coefficient SCC based on the theoretical temperature value P and the actual temperature value R; 40) Evaluate the wind turbine based on the temperature deviation DEV, temperature mean square error MSE and correlation coefficient SCC health status. With the help of the above fault early warning method, the precision and accuracy of evaluating the health state of the wind power generating set can be improved.

Description

The fault early warning method of wind power generating set

Technical field

The present invention relates to wind generating technology, relate in particular to a kind of fault early warning method that is applicable to wind power generating set.

Background technology

Along with the consumption of the nonrenewable resources such as oil, coal and the raising of environmental consciousness, people more and more pay attention to the application of regenerative resource.Wherein, wind energy is as a kind of clean regenerative resource, and the status in global energy-saving and emission-reduction work becomes more and more important, and correspondingly, the wind power generating set that wind energy can be converted to electric energy is also more and more paid attention to and develops.

Conventionally, wind power generating set is all installed on field, and its working environment is often more severe, thereby easily breaks down.In order to find in time the fault of wind power generating set, produce the problems such as generating efficiency reduction, original paper damage with the wind power generating set that prevents operation exception, conventionally can monitor the operation conditions of wind power generating set.Because the temperature conditions of wind power generating set often can reflect the operation conditions of wind power generating set, thereby often adopt in actual applications the method for monitoring wind power generating set temperature to monitor the operation conditions of wind power generating set.

Refer to Fig. 1, wherein show a kind of conventional temperature of passing through monitoring wind power generating set and monitor the method for wind power generating set operation conditions.In the method, first gather the temperature data such as the point being monitored such as pitch-controlled system and/or current transformer, then gathered temperature data and predefined associated temperature threshold value are compared, if the temperature data gathering is greater than predefined associated temperature threshold value, report to the police to point out wind power generating set operation exception; Otherwise, represent this wind power generating set normal operation, and method ends flow process.

Although method is widely adopted at present shown in Fig. 1, still inevitably there is following problems in it in actual applications:

One; owing to only having set a temperature threshold corresponding to each point being monitored; thereby; whether method shown in Fig. 1 can only exceed this temperature threshold based on the current temperature in point being monitored, and to judge the running status of wind power generating set be normal or abnormal; that is to say; method shown in Fig. 1 can only be divided into the running status of wind power generating set normally and abnormal two kinds, and cannot divide more subtly, thereby the assessment degree of accuracy of method shown in Fig. 1 is lower.In this case, due in the time that the running status of wind power generating set levels off to abnormality, still judge that by method shown in Fig. 1 the running status of current wind power generating set is normal, cause finding in time the fault of wind power generating set, thereby the wind power generating set of convergence abnormality is at high temperature worked for a long time, so that produce the problems such as shortening in serviceable life, generating efficiency reduction.In addition; to depart from normal operating condition far away or depart from slightly normal condition owing to cannot assess wind power generating set by method shown in Fig. 1; if while causing wind power generating set still can keep normally moving in departing from normal condition slightly; still can report to the police; thereby make operating personnel or automatic control system carry out shutdown operation to overhaul based on this warning; cause the waste of manpower and material resources cost, and have a strong impact on the generating efficiency of wind power generating set.

They are two years old, owing to only having set a temperature threshold corresponding to each point being monitored, thereby, in the time that this temperature threshold itself arranges to such an extent that have the temperature data of problem and/or collection wrong, still the running status that relies on the comparative result between gathered temperature and temperature threshold and pass judgment on wind power generating set tends to cause erroneous judgement, thereby cause wrong report or fail to report, and therefore and impact subsequent operation measure that wind power generating set is implemented, for example, while occurring reporting by mistake when wind power generating set script normal operation, operating personnel or automatic control system will make wind power generating set shut down to overhaul based on this warning, this will cause the waste of manpower and material resources cost, and have a strong impact on the generating efficiency of wind power generating set.And in the situation that failing to report, wind power generating set can be operated at abnormal temperature and the final fault that produces for a long time, like this, not only can be because shortening the serviceable life of blower fan, but also can have a strong impact on because of shutdown maintenance the generating efficiency of wind power generating set.

Summary of the invention

For solving the problems of the technologies described above, the invention provides a kind of fault early warning method of wind power generating set, health status degree of accuracy and the accuracy of its assessment wind power generating set are higher.

For this reason, the invention provides the fault early warning method of wind power generating set, it comprises the steps:

10) within the sampling period, gather the actual temperature value R of N point being monitored and the supplemental characteristic with temperature correlation;

20) input predetermined temperature model by the actual temperature value R of gathered point being monitored and with the supplemental characteristic of temperature correlation, calculate the temperature prediction value of point being monitored, and theoretical temperatures value P using described temperature prediction value as point being monitored;

30) based on theoretical temperatures value P and actual temperature value R and obtain the two temperature deviation DEV, temperature mean square deviation MSE and related coefficient SCC, wherein,

DEV=P-R

$\sqrt{\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(P\left(i\right)-R\left(i\right))}^2}$

$\mathrm{SCC}=\frac{N\underset{i=1}{\overset{N}{\mathrm{\Σ}}}P\left(i\right)R\left(i\right)-\underset{i=1}{\overset{N}{\mathrm{\Σ}}}P\left(i\right)\underset{i=1}{\overset{N}{\mathrm{\Σ}}}R\left(i\right)}{\sqrt{N\underset{i=1}{\overset{N}{\mathrm{\Σ}}}P{\left(i\right)}^2-{\left(\underset{i=1}{\overset{N}{\mathrm{\Σ}}}P\left(i\right)\right)}^2}\sqrt{N\underset{i=1}{\overset{N}{\mathrm{\Σ}}}R{\left(i\right)}^2-{\left(\underset{i=1}{\overset{N}{\mathrm{\Σ}}}R\left(i\right)\right)}^2}}$

N is the sampling total degree in sampling period, and i is sampling sequence number, i=1, and 2 ..., N;

40) based on temperature deviation DEV, temperature mean square deviation MSE and related coefficient SCC and assess the health status of wind power generating set.

Preferably, in described step 10) and 20) between also comprise step 15): compare with corresponding setting threshold respectively by the actual temperature R of gathered point being monitored and with the supplemental characteristic of temperature correlation, if the former exceeds the normal range that the latter limits, report to the police and method ends flow process; If the former does not exceed the normal range that the latter limits, enter follow-up evaluation process.

Wherein, described point being monitored comprises the point being monitored being positioned in the individual different pitch motor of n.

Wherein, in step 15), if the former does not exceed the normal range that the latter limits, calculate the temperature-averaging value T of the point being monitored of n different pitch motor, and calculate the medial temperature difference SUB of two pitch motors in the pitch motor that n is different, and wherein, SUB=T _j-T _r, j ≠ r, j=1,2 ..., n, r=1,2 ..., n; Then the temperature gap SUB of pitch motor and the threshold value of setting are compared, if the former is greater than setting threshold, reports to the police and method ends flow process, otherwise forward step 20 to).

Wherein, described point being monitored comprises the point being monitored being positioned on current transformer.

Wherein, in step 40) in, described temperature mean square deviation MSE, related coefficient SCC and temperature deviation DEV are compared with corresponding threshold value respectively, and judge whether the former exceeds corresponding threshold value, then the health status of the health status level evaluation wind power generating set based on dividing in advance.

Wherein, in described health status grade of dividing in advance, the threshold interval of m different range is set with respect to described temperature deviation DEV, to obtain the health status grade after further refinement, wherein, m is natural number;

In step 40) in, judge the threshold interval at temperature deviation DEV place, and add up the quantity of temperature deviation DEV in each threshold interval, the then health status of the health status level evaluation wind power generating set based on after further refinement.

Wherein, obtain like this health status grade after described further refinement, that is, with respect to temperature mean square deviation MSE, related coefficient SCC, mean square deviation threshold value MS, correlation coefficient threshold SS are set respectively; With respect to temperature deviation DEV, 3 temperature deviation threshold values are set, be respectively 1S, 2S and 3S, wherein, 1S < 2S < 3S, and mark off from small to large 4 threshold intervals by described 3 temperature deviation threshold values, be respectively (∞, 1S], (1S, 2S], (2S, 3S] and (3S ,+∞); And based on described mean square deviation threshold value MS, correlation coefficient threshold SS and 4 threshold intervals, the health status grade of wind power generating set is divided into following 9 grades by excellent to poor:

Grade 1: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, related coefficient SCC is lower than correlation coefficient threshold SS, and the temperature deviation DEV of the N secondary data gathering all in threshold interval (∞, 1S];

Grade 2: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV that has 1 secondary data in gathered N secondary data in threshold interval (1S, 2S], the temperature deviation DEV of all the other N-1 secondary data all in threshold interval (∞, 1S];

Grade 3: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV that has 2 secondary data in gathered N secondary data in threshold interval (1S, 2S], and the collection behavior of 2 described secondary data is discontinuous, the temperature deviation DEV of all the other N-2 secondary data all in threshold interval (∞, 1S];

Class 4: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV that has 1 secondary data in gathered N secondary data in threshold interval (2S, 3S], the temperature deviation DEV of all the other N-1 secondary data all in threshold interval (∞, 1S];

Class 5: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV that has 2 secondary data in gathered N secondary data in threshold interval (1S, 2S], and the collection behavior of 2 described secondary data is continuous, the temperature deviation DEV of all the other N-2 secondary data all in threshold interval (∞, 1S];

Class 6: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV that has 3 secondary data in gathered N secondary data in threshold interval (1S, 2S], the temperature deviation DEV of all the other N-3 secondary data all in threshold interval (∞, 1S];

Grade 7: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV that has 4 secondary data in gathered N secondary data in threshold interval (1S, 2S], the temperature deviation DEV of all the other N-4 secondary data all in threshold interval (∞, 1S];

Grade 8: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And in gathered N secondary data, have 2 times or the temperature deviation DEV of 3 secondary data in threshold interval (2S, 3S], the temperature deviation DEV of remainder data in threshold interval (∞, 1S];

Grade 9: temperature mean square deviation MSE has the temperature deviation DEV of 1 secondary data in threshold interval (3S higher than mean square deviation threshold value MS, related coefficient SCC higher than correlation coefficient threshold SS, in gathered N secondary data, + ∞), and/or in the N secondary data gathering, has the temperature deviation DEV of 4 secondary data at least in threshold interval (2S, 3S], the temperature deviation DEV of remainder data in threshold interval (∞, 1S].

Wherein, in described step 40) in, determine the health status grade of wind power generating set based on temperature deviation DEV, temperature mean square deviation MSE and related coefficient SCC, and

When the health status grade of wind power generating set during to grade 3, judges that the running status of current wind power generating set is as normal in grade 1, and method ends flow process; When the health status grade of wind power generating set, is judged that the running status of current wind power generating set is as departing from normal or being tending towards abnormal, and given a warning during to class 6 in class 4; When the health status grade of wind power generating set, is judged that the running status of current wind power generating set is as abnormal, and given the alarm during to grade 9 in grade 7.

Preferably, in described step 10) afterwards, store the actual temperature value R of the point being monitored gathering and the supplemental characteristic with temperature correlation.

Wherein, described and supplemental characteristic temperature correlation comprises power, voltage, electric current, wind speed and the environment temperature of wind power generating set.

The present invention has following beneficial effect:

The fault early warning method of wind power generating set provided by the invention, it is by setting up in advance temperature model, the actual temperature value R of N point being monitored that can be based on gathering within the sampling period and obtain the temperature prediction value of gathered point being monitored with the supplemental characteristic of temperature correlation, in order to the temperature theoretical value P as point being monitored; Then, based on temperature theoretical value P and actual temperature value R and obtain the two temperature deviation DEV, temperature mean square deviation MSE and related coefficient SCC, and based on temperature deviation DEV, temperature mean square deviation MSE and related coefficient SCC and assess the health status of wind power generating set.By means of above-mentioned fault early warning method, can be based on temperature deviation DEV, temperature mean square deviation MSE and related coefficient SCC and set multiple different temperature thresholds, this not only can improve the accuracy of set temperature threshold value, thereby improve the accuracy of the health status of assessment wind power generating set, and then reduce the wrong report of alarm or the situation about failing to report even avoided; And, health status grade that can further refinement wind power generating set, thus the degree of accuracy of the health status of assessment wind power generating set improved, and then reduce manpower and material resources cost, improve the generating efficiency of wind power generating set.

Accompanying drawing explanation

Fig. 1 is the FB(flow block) of the method for the monitoring wind power generating set operation conditions commonly used;

Fig. 2 is the FB(flow block) of the fault early warning method of wind power generating set provided by the invention; And

Fig. 3 is the FB(flow block) of the fault early warning method of the pitch motor of wind power generating set provided by the invention.

Embodiment

For making those skilled in the art person understand better technical scheme of the present invention, below in conjunction with accompanying drawing, the fault early warning method of wind power generating set provided by the invention is elaborated.

Fig. 2 is the FB(flow block) of the fault early warning method of wind power generating set provided by the invention.Refer to Fig. 2, the fault early warning method of wind power generating set comprises the following steps:

Step S10 gathers the actual temperature value R of N point being monitored and the supplemental characteristic with temperature correlation within the sampling period.Wherein, sampling period and sampling number can be set according to actual conditions, for example, a sampling period can be set as to 2 hours, and gather the actual temperature value R of 1 wind power generating set and the supplemental characteristic with temperature correlation, gather altogether 12 times for every 10 minutes in 2 hours.

Power, voltage, electric current, wind speed and the environment temperatures etc. that in actual applications, can comprise wind power generating set with the supplemental characteristic of temperature correlation and the temperature of wind power generating set have the data of direct or indirect relation.

Step S20, inputs predetermined temperature model by the actual temperature value R of gathered point being monitored and with the supplemental characteristic of temperature correlation, calculates the temperature prediction value of point being monitored, and theoretical temperatures value P using this temperature prediction value as point being monitored.The method for building up of temperature model is specially: first obtain and the variable of temperature correlation that affects wind power generating set by empirical analysis and data mining; Then, the historical data relevant to variable while utilizing the wind power generating set that gathers normally to move, the modeling method of employing artificial intelligence is set up temperature model.By temperature model, the temperature prediction value that obtains point being monitored with the supplemental characteristic of temperature correlation of point being monitored that can be based on gathered, and set it as theoretical temperatures value P.

Step S30, based on theoretical temperatures value P and actual temperature value R and obtain the two temperature deviation DEV, temperature mean square deviation MSE and related coefficient SCC, wherein,

DEV＝P_DATA-R_DATA

$\mathrm{MSE}=\sqrt{\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}{(P\_\mathrm{DATA}\left(i\right)-R\_\mathrm{DATA}\left(i\right))}^2}$

$\mathrm{SCC}=\frac{N\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}P\_\mathrm{DATA}\left(i\right)R\_\mathrm{DATA}\left(i\right)-\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}P\_\mathrm{DATA}\left(i\right)\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}R\_\mathrm{DATA}\left(i\right)}{\sqrt{N\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}P\_\mathrm{DATA}{\left(i\right)}^2-{(\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}P\_\mathrm{DATA}\left(i\right))}^2}\sqrt{N\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}R\_\mathrm{DATA}{\left(i\right)}^2-{(\underset{i=1}{\overset{N}{\mathrm{\&Sigma;}}}R\_\mathrm{DATA}\left(i\right))}^2}}$

N is the sampling total degree in sampling period, and i is sampling sequence number, i=1, and 2 ..., N.

Step S40, based on temperature deviation DEV, temperature mean square deviation MSE and related coefficient SCC and assess the health status of wind power generating set.Particularly, first, divide in advance the health status grade of wind power generating set, that is, with respect to temperature mean square deviation MSE and related coefficient SCC, mean square deviation threshold value and correlation coefficient threshold are set respectively; The threshold interval of m different range is set with respect to temperature deviation DEV, that is, m-1 different temperature deviation threshold value is set, can mark off m threshold interval by m-1 different temperature deviation threshold value.By the threshold interval of multiple different range is set with respect to temperature deviation DEV, not only can improve the accuracy of the health status of assessment wind power generating set, and health status grade that can further refinement wind power generating set, thereby improve the accuracy of the health status of assessment wind power generating set.

Below by how giving an example to illustrate based on temperature deviation DEV, temperature mean square deviation MSE and related coefficient SCC and divide the health status grade of wind power generating set.

Particularly, first, with respect to temperature mean square deviation MSE, related coefficient SCC, mean square deviation threshold value MS, correlation coefficient threshold SS are set respectively, with respect to temperature deviation DEV, 3 temperature deviation threshold values are set, be respectively 1S, 2S and 3S, wherein, 1S < 2S < 3S, and, mark off from small to large 4 threshold intervals by these 3 temperature deviation threshold values, be respectively (∞, 1S], (1S, 2S], (2S, 3S] and (3S, + ∞), wherein, threshold interval (∞, 1S], (1S, 2S], (2S, 3S] be a left side and open the interval that the right side is closed, , get temperature deviation threshold value 1S, 2S and 3S are as threshold interval (∞, 1S], (1S, 2S], (2S, 3S] upper limit numerical value.

Based on mean square deviation threshold value MS, correlation coefficient threshold SS and 4 threshold intervals, the health status grade of wind power generating set is divided into 9 grades by excellent to poor, the concrete Rules of Assessment of the health status based on these 9 level evaluation wind power generating set is as follows:

Grade 1: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV of the N secondary data gathering all in threshold interval (∞, 1S];

Grade 2: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV that has 1 secondary data in gathered N secondary data in threshold interval (1S, 2S], the temperature deviation DEV of all the other N-1 secondary data all in threshold interval (∞, 1S];

Grade 3: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV that has 2 secondary data in gathered N secondary data in threshold interval (1S, 2S], and the collection behavior of 2 described secondary data is discontinuous, the temperature deviation DEV of all the other N-2 secondary data all in threshold interval (∞, 1S].The collection behavior of so-called 2 secondary data is discontinuous, refers to that the order of 2 image data is non-conterminous.For example, the sampling sequence number that has 1 secondary data in 2 secondary data is 2, i.e. the 2nd image data, if the sampling sequence number of another secondary data is 1 or 3, the collection behavior of 2 secondary data is continuous; If the sampling sequence number of another secondary data is other sampling sequence number except 1 or 3, the collection behavior of 2 secondary data is discontinuous.

Class 4: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV that has 1 secondary data in gathered N secondary data in threshold interval (2S, 3S], the temperature deviation DEV of all the other N-1 secondary data all in threshold interval (∞, 1S];

Class 5: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV that has 2 secondary data in gathered N secondary data in threshold interval (1S, 2S], and the collection behavior of 2 secondary data is continuous, the temperature deviation DEV of all the other N-2 secondary data all in threshold interval (∞, 1S];

Class 6: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV that has 3 secondary data in gathered N secondary data in threshold interval (1S, 2S], the temperature deviation DEV of all the other N-3 secondary data all in threshold interval (∞, 1S];

Grade 7: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And the temperature deviation DEV that has 4 secondary data in gathered N secondary data in threshold interval (1S, 2S], the temperature deviation DEV of all the other N-4 secondary data all in threshold interval (∞, 1S];

Grade 8: temperature mean square deviation MSE is lower than mean square deviation threshold value MS, and related coefficient SCC is lower than correlation coefficient threshold SS; And in gathered N secondary data, have 2 times or the temperature deviation DEV of 3 secondary data in threshold interval (2S, 3S], the temperature deviation DEV of remainder data in threshold interval (∞, 1S];

Grade 9: temperature mean square deviation MSE has the temperature deviation DEV of 1 secondary data in threshold interval (3S higher than mean square deviation threshold value MS, related coefficient SCC higher than correlation coefficient threshold SS, in gathered N secondary data, + ∞), and/or in the N secondary data gathering, has the temperature deviation DEV of 4 secondary data at least in threshold interval (2S, 3S], the temperature deviation DEV of remainder data in threshold interval (∞, 1S].

In the time that the appraisal procedure that adopts above-mentioned health status is carried out fault pre-alarming to the running status of wind power generating set, when the health status grade of wind power generating set is in grade 1 during to grade 3, judge that the running status of current wind power generating set is as normal, and method ends flow process; When the health status grade of wind power generating set, is judged that the running status of current wind power generating set is as departing from normal or being tending towards abnormal, and given a warning during to class 6 in class 4; When the health status grade of wind power generating set, is judged that the running status of current wind power generating set is as abnormal, and given the alarm during to grade 9 in grade 7.

It should be noted that, it is more than an object lesson dividing the health status grade of wind power generating set, but the present invention is not limited thereto, in actual applications, size that can be based on temperature deviation DEV and the size of Changing Pattern, temperature mean square deviation MSE and related coefficient SCC, and adopt the method for data mining to divide the health status grade of wind power generating set.

Be described in detail for the workflow of the health status of assessing wind power generating set below.

Particularly, first, temperature deviation DEV is compared with the threshold interval of m the different range setting in advance, temperature mean square deviation MSE and mean square deviation threshold value are compared and related coefficient SCC and correlation coefficient threshold are compared; Then, judge whether temperature mean square deviation MSE exceeds mean square deviation threshold value, and whether related coefficient SCC exceeds correlation coefficient threshold, meanwhile, judge the quantity of temperature deviation DEV in which threshold interval and the temperature deviation DEV in each threshold interval; Finally, the health status of the health status level evaluation wind power generating set of the wind power generating set based on dividing in advance.

In addition, after step S10, can store to the actual temperature value R of gathered point being monitored and with the supplemental characteristic of temperature correlation.

In order to contribute to those skilled in the art to understand further technical scheme of the present invention, take the health status of pitch motor of assessment wind power generating set as example, further the fault early warning method of wind power generating set provided by the invention is described in detail below.

The FB(flow block) of the fault early warning method of the pitch motor that particularly, Fig. 3 is wind power generating set provided by the invention.Refer to Fig. 3, in the method, point being monitored comprises that n is positioned at the point being monitored in different pitch motors, that is, the quantity of pitch motor is n, and a point being monitored is set in each pitch motor.

Step S100 gathers the actual temperature value R of N point being monitored and the supplemental characteristic with temperature correlation within the sampling period.

Step S200, compares with corresponding setting threshold by the actual temperature value R of gathered point being monitored and with the supplemental characteristic of temperature correlation.In actual applications, above-mentioned setting threshold can be set to the mxm. that allows the data that gather to reach, i.e. secure threshold.

Step S300, judges whether the former exceeds the normal range that the latter limits, and if so, sends the warning the method ends flow process that represent severely subnormal; If not, flow process enters step S400.

Step S400, calculates the temperature-averaging value T of the point being monitored of n different pitch motor, and calculates the medial temperature difference SUB of two pitch motors in n different pitch motor, wherein, and SUB=T _j-T _r, j ≠ r, j=1,2 ... n, r=1,2 ... n.

Step S500, compares the temperature gap SUB of pitch motor and the threshold value of setting.

Step S600, judges whether the former exceeds the normal range that the latter limits, and if so, reports to the police and method ends flow process; If not, flow process enters step S700.

Step S700, enters the workflow of health status of the pitch motor of assessment wind power generating set.Because the fault early warning method of wind power generating set shown in the workflow of health status of the pitch motor of assessment wind power generating set and Fig. 2 is similar, do not repeat them here.

It should be noted that, in actual applications, the fault early warning method of the pitch motor of wind power generating set also can save step S300, and/or save step S400, step S500 and step S600,, after completing steps S100, flow process directly enters step S700, the workflow of the health status of the pitch motor of assessment wind power generating set.

Take the health status of current transformer of assessment wind power generating set as example, further the fault early warning method of wind power generating set provided by the invention is described in detail below.Particularly, compared with the fault early warning method of the pitch motor of wind power generating set shown in Fig. 3, the fault early warning method of the current transformer of wind power generating set comprises step S100, step S200, step S300 and step S700 equally, its difference is: a wind power generating set is only provided with a current transformer, and point being monitored only comprises the point being monitored being positioned on current transformer.Therefore, the fault early warning method of the current transformer of wind power generating set does not need to calculate the temperature-averaging value of point being monitored and temperature-averaging value is carried out to corresponding poor comparison, thereby has saved above-mentioned steps S400, step S500 and step S600.

In sum, the fault early warning method of wind power generating set provided by the invention, it is by setting up in advance temperature model, the actual temperature value R of N point being monitored that can be based on gathering within the sampling period and obtain the temperature prediction value of gathered point being monitored with the supplemental characteristic of temperature correlation, in order to the temperature theoretical value P as point being monitored; Then, based on temperature theoretical value P and actual temperature value R and obtain the two temperature deviation DEV, temperature mean square deviation MSE and related coefficient SCC, and based on temperature deviation DEV, temperature mean square deviation MSE and related coefficient SCC and assess the health status of wind power generating set.By means of above-mentioned fault early warning method, can be based on temperature deviation DEV, temperature mean square deviation MSE and related coefficient SCC and set multiple different temperature thresholds, this not only can improve the accuracy of set temperature threshold value, thereby improve the accuracy of the health status of assessment wind power generating set, and then reduce the wrong report of alarm or the situation about failing to report even avoided; And, health status grade that can further refinement wind power generating set, thus the degree of accuracy of the health status of assessment wind power generating set improved, and then reduce manpower and material resources cost, improve the generating efficiency of wind power generating set.

Be understandable that, above embodiment is only used to principle of the present invention is described and the illustrative embodiments that adopts, but the present invention is not limited thereto.For those skilled in the art, without departing from the spirit and substance in the present invention, can make various modification and improvement, these modification and improvement are also considered as protection scope of the present invention.

Claims (11)

1. A fault early warning method for a wind power generating set, characterized in that it may further comprise the steps: 10) Collect the actual temperature value R and temperature-related parameter data of the monitored point N times within the sampling period; 20) Input the collected actual temperature value R of the monitored point and temperature-related parameter data into the predetermined temperature model, calculate the temperature prediction value of the monitored point, and use the temperature prediction value as the monitored point Theoretical temperature value P; 30) Based on the theoretical temperature value P and the actual temperature value R, the temperature deviation DEV, temperature mean square error MSE and correlation coefficient SCC are obtained, where, DEV=P-R $\sqrt{\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; N</span>}}\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}$ $\mathrm{<span\; class="notranslate">\; SCC</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; N</span>}\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}{\sqrt{\mathrm{<span\; class="notranslate">\; N</span>}\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; P</span>}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{<span\; class="notranslate">\; (</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}\sqrt{\mathrm{<span\; class="notranslate">\; N</span>}\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; R</span>}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span>{<span\; class="notranslate">\; (</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; R</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}}$ N is the total number of samples in the sampling period, i is the sampling number, i=1,2,...,N; 40) Evaluate the health status of wind turbines based on temperature deviation DEV, temperature mean square error MSE and correlation coefficient SCC. 2. The fault early warning method for wind power generating sets according to claim 1, characterized in that, between the steps 10) and 20), a step 15) is further included: the collected actual temperature R of the monitored point and The temperature-related parameter data are compared with the corresponding set thresholds. If the former exceeds the normal range defined by the latter, an alarm will be issued and the process of the method will end; if the former does not exceed the normal range defined by the latter, the subsequent evaluation process. 3 . The fault early warning method for a wind power generating set according to claim 2 , wherein the monitored points include monitored points located on n different pitch motors. 4 . 4. The fault early warning method for wind power generating sets according to claim 3, characterized in that, in step 15), if the former does not exceed the normal range defined by the latter, then calculate the n different pitch motors by The temperature average T of the monitoring point, and calculate the average temperature difference SUB of two pitch motors among n different pitch motors, where, SUB=T _j -T _r , j≠r, j=1,2, ...,n, r=1,2,...,n; Then compare the temperature difference SUB of the pitch motor with the set threshold, if the former is greater than the set threshold, then alarm and end the process of the method, otherwise go to step 20). 5 . The fault early warning method for a wind power generating set according to claim 2 , wherein the monitored point includes a monitored point located on a converter. 6 . 6. The fault early warning method for wind power generators according to claim 1, characterized in that, in step 40), the temperature mean square error MSE, correlation coefficient SCC and temperature deviation DEV are compared with corresponding thresholds, And judge whether the former exceeds the corresponding threshold, and then evaluate the health status of the wind turbine based on the pre-divided health status level. 7. The fault early warning method for wind power generators according to claim 6, characterized in that, in the pre-divided health status levels, m threshold intervals of different ranges are set with respect to the temperature deviation DEV, to obtain The health status level after further refinement, where m is a natural number; In step 40), the threshold interval of the temperature deviation DEV is determined, and the number of temperature deviation DEV in each threshold interval is counted, and then the health status of the wind turbine is evaluated based on the further refined health status level. 8. The fault early warning method for wind power generators according to claim 7, characterized in that the further refined health status level is obtained in this way, that is, the mean square error is respectively set with respect to the temperature mean square error MSE and the correlation coefficient SCC Threshold MS, correlation coefficient threshold SS; set three temperature deviation thresholds relative to temperature deviation DEV, which are 1S, 2S and 3S respectively, where 1S<2S<3S, and are divided by the three temperature deviation thresholds from small to large There are 4 threshold intervals, namely (-∞, 1S], (1S, 2S], (2S, 3S] and (3S, +∞); and based on the mean square error threshold MS, correlation coefficient threshold SS and 4 The health status of wind turbines is divided into the following 9 levels from excellent to poor: Level 1: The temperature mean square error MSE is lower than the mean square error threshold MS, the correlation coefficient SCC is lower than the correlation coefficient threshold SS, and the temperature deviation DEV of the collected N data is in the threshold interval (-∞, 1S]; Level 2: The temperature mean square error MSE is lower than the mean square error threshold MS, and the correlation coefficient SCC is lower than the correlation coefficient threshold SS; and the temperature deviation DEV of one of the N data collected is in the threshold interval (1S, 2S], The temperature deviation DEV of the remaining N-1 data are all in the threshold interval (-∞, 1S]; Level 3: The temperature mean square error MSE is lower than the mean square error threshold MS, and the correlation coefficient SCC is lower than the correlation coefficient threshold SS; and the temperature deviation DEV of 2 times of the collected N data is in the threshold interval (1S, 2S], And the collection behavior of the 2 times of data is discontinuous, and the temperature deviation DEV of the remaining N-2 times of data are all in the threshold interval (-∞, 1S]; Level 4: The temperature mean square error MSE is lower than the mean square error threshold MS, and the correlation coefficient SCC is lower than the correlation coefficient threshold SS; and the temperature deviation DEV of the data in one of the N times of data collected is in the threshold interval (2S, 3S], The temperature deviation DEV of the remaining N-1 data are all in the threshold interval (-∞, 1S]; Level 5: The temperature mean square error MSE is lower than the mean square error threshold MS, and the correlation coefficient SCC is lower than the correlation coefficient threshold SS; and the temperature deviation DEV of 2 data in the N times of data collected is in the threshold interval (1S, 2S], And the collection behavior of the 2 times of data is continuous, and the temperature deviation DEV of the remaining N-2 times of data are all in the threshold interval (-∞, 1S]; Level 6: The temperature mean square error MSE is lower than the mean square error threshold MS, and the correlation coefficient SCC is lower than the correlation coefficient threshold SS; and the temperature deviation DEV of 3 times of the collected N data is in the threshold interval (1S, 2S], The temperature deviation DEV of the remaining N-3 times of data are all in the threshold interval (-∞, 1S]; Level 7: The temperature mean square error MSE is lower than the mean square error threshold MS, and the correlation coefficient SCC is lower than the correlation coefficient threshold SS; and the temperature deviation DEV of 4 times of the collected N data is in the threshold interval (1S, 2S], The temperature deviation DEV of the remaining N-4 times of data are all in the threshold interval (-∞, 1S]; Level 8: The temperature mean square error MSE is lower than the mean square error threshold MS, and the correlation coefficient SCC is lower than the correlation coefficient threshold SS; and the temperature deviation DEV of 2 or 3 times of the collected N data is in the threshold interval (2S, 3S], the temperature deviation DEV of the rest of the data is in the threshold interval (-∞, 1S]; Level 9: The temperature mean square error MSE is higher than the mean square error threshold MS, the correlation coefficient SCC is higher than the correlation coefficient threshold SS, and the temperature deviation DEV of 1 data in the N times of data collected is in the threshold interval (3S, +∞), And/or the temperature deviation DEV of at least 4 of the collected N data is in the threshold interval (2S, 3S], and the temperature deviation DEV of the rest of the data is in the threshold interval (-∞, 1S]. 9. The fault early warning method for a wind power generating set according to claim 8, characterized in that, in the step 40), the health status of the wind generating set is determined based on the temperature deviation DEV, the temperature mean square error MSE and the correlation coefficient SCC grade, and When the health state level of the wind power generating set is in level 1 to level 3, it is determined that the current operating state of the wind power generating set is normal, and the process of this method is ended; when the health state level of the wind power generating set is in level 4 to level 6, it is determined The current operating state of the wind power generating set deviates from normal or tends to be abnormal, and a warning is issued; when the health status of the wind generating set is at level 7 to level 9, it is determined that the current operating state of the wind generating set is abnormal, and an alarm is issued. 10 . The fault early warning method for wind power generating sets according to claim 1 , characterized in that, after the step 10), the collected actual temperature value R of the monitored point and parameter data related to temperature are stored. 11 . 11. The fault warning method for a wind power generating set according to any one of claims 1-10, wherein the temperature-related parameter data includes power, voltage, current, wind speed and ambient temperature of the wind generating set.

Images