阅读说明：本技术 一种风力发电机测风仪健康状态检测方法 (Health state detection method for wind meter of wind driven generator ) 是由 李永军 刘军 于 2021-07-29 设计创作，主要内容包括：本发明涉及风力发电机检测领域,特别涉及一种风力发电机测风仪健康状态检测方法,包括以下步骤：S1、获取SCADA数据,并对数据进行过滤；S2、对过滤后的数据进行预处理；S3、构建测风仪健康评分指标及判别标准；S4、根据采集的数据分析测风仪健康状态。本发明的有益效果在于：本发明对风机测风仪的健康状况监测成本低、潜力大,充分利用SCADA监控数据,实现检测的智能化、自动化、实时性,除此之外,本发明提供数据预处理方案保证了测风仪风速和测风塔风速的相关性,使二者关系更加清晰呈现,同时本发明的思路可以延申到更多的风机设备组件的检测当中,对风机整个的设备检测存在重大意义。(The invention relates to the field of detection of wind driven generators, in particular to a method for detecting the health state of a wind meter of a wind driven generator, which comprises the following steps: s1, obtaining SCADA data and filtering the data; s2, preprocessing the filtered data; s3, constructing a health scoring index and a judgment standard of the anemometer; and S4, analyzing the health state of the anemometer according to the collected data. The invention has the beneficial effects that: the invention has low cost and great potential for monitoring the health condition of the wind meter of the fan, fully utilizes SCADA monitoring data, realizes the intellectualization, automation and real-time performance of detection, in addition, the invention provides a data preprocessing scheme to ensure the correlation between the wind speed of the wind meter and the wind speed of the wind measuring tower, so that the correlation between the wind speed and the wind speed of the wind measuring tower is more clearly shown, meanwhile, the idea of the invention can be extended to the detection of more fan equipment components, and has great significance for the whole equipment detection of the fan.)

1. A health state detection method for a wind meter of a wind driven generator is characterized by comprising the following steps:

s1, obtaining SCADA data and filtering the data;

s2, preprocessing the filtered data;

s3, constructing a health scoring index and a judgment standard of the anemometer;

and S4, analyzing the health state of the anemometer according to the collected data.

2. The method for detecting the health state of the wind meter of the wind driven generator according to claim 1, wherein the step S1 is specifically to remove data which cannot well reflect the real state of the wind turbine, including data of a wind turbine generator shutdown and a power-limiting operation state, and data of a wind speed lower than a first threshold value and a power generation power lower than a second threshold value.

3. The method for detecting the health state of the anemometer of the wind turbine according to claim 1, wherein the step S2 specifically comprises:

s2.1, acquiring the maximum wind speed V in the wind speed data_maxAnd minimum wind speed V_min；

S2.2, dividing the SCADA data into N intervals, wherein

E is a set speed; if the calculated N is a non-integer, taking the minimum positive integer larger than the calculated value;

s2.3, defining a numerical value K, wherein the initial value of the K is 1;

s2.4, calculating a wind speed range and a wind speed expected value in the Kth interval; calculating the expected power value of the generator in the Kth interval;

s2.5, determining that K is greater than N, if not, repeating step S2.4 after K is equal to K + 1; thus, the relation between the generator power and the wind speed is obtained.

4. The method for detecting the health state of the anemometer of the wind turbine according to claim 3, wherein the step S2.4 specifically comprises:

s2.41, acquiring the wind speed range of the Kth wind speed interval,

[V_min+(K-1)V_S，V_min+KV_S]wherein

S2.42, calculating the expected wind speed value in the Kth interval:

in the formula v_iIs the wind speed; m is the number of different wind speed data in the Kth interval; p is a radical of_jIs the wind speed v_jThe probability of the value of (b) appearing in the Kth interval, whereinn is the number of data in the Kth interval, n_jThe wind speed in the Kth interval is v_jThe number of data of (2);

s2.43, calculating the expected value of the power of the generator in the Kth interval:

in the formula: p_iIs a power value; q is the number of different power data in the Kth interval; o_iIs a power P_iThe probability that the corresponding wind speed occurs in the kth interval,r is the number of power data in the K-th interval, r_iFor the power in the Kth intervalIs P_iThe number of data of (2).

5. The method for detecting the health state of the anemometer of the wind turbine according to claim 1, wherein the step S3 specifically comprises:

s3.1, constructing a relation function of anemometer measurement values and anemometer tower anemometer values;

s3.2, establishing a health scoring index of the anemometer according to the relation function in the S3.1;

and S3.3, constructing a judgment standard of the health state of the anemometer.

6. The method for detecting the health state of the wind meter of the wind driven generator according to claim 5, wherein the step S3.1 is specifically as follows:

constructing a relation function of a anemometer measurement value X and a anemometer tower anemometer value Y,

SCADA data { x) for describing any set of inputs_i，y_iThe relationship between (i ═ 1, 2, … …, g); wherein a is_j(j-0, 1, 2, … …, z) are model coefficients,is y_iAn estimate of (d).

7. The method for detecting the health status of the anemometer of the wind turbine as claimed in claim 6, wherein the step S3.1 further comprises introducing a matrix

Wherein R is²Is the sum of the squares of the errors,

by substituting formula (6) into (5), A ═ X^TY(X^TX)^-1 (8)；

Calculating a from the above equation₀～a_zIs determined, input SCADA data { x_i，y_iThe relationship between them.

8. The method for detecting the health state of the anemometer of the wind turbine according to claim 5, wherein the step S3.2 specifically comprises:

constructing a health scoring index S of the anemometer,

in the formula: a is_jModel coefficients calculated from the current SCADA data; b_jCalculating a model coefficient obtained from historical SCADA data of the fan in normal operation; x_maxAnd X_minThe maximum value and the minimum value in X are respectively, and K is the K-th interval.

9. The method for detecting the health state of the wind meter of the wind driven generator according to claim 5, wherein a discrimination standard is established:

s is 0, the fan anemoscope runs normally;

s >0 indicates that the wind meter of the fan has faults, and the faults are more serious when the numerical value is larger.

10. The method for detecting the health state of the wind meter of the wind driven generator according to claim 5, further comprising a step S3.4, specifically:

constructing fault classification according to corresponding deviation between the generated energy calculated by the wind speed value and the actual generated energy, and finally setting the generated energy deviation caused by the wind speed deviation caused by the fault as P_offsetActual power generation amount is P_cap；

When in useWhen S is 0, the fault grade is I, and the severity is normal;

when in useWhen S is 0, the fault grade is II, and the severity is slight;

when in useWhen S is 0, the fault grade is III, and the severity is severe;

when in useThe failure grade is IV when S is 0, and the severity is very serious.

Technical Field

The invention relates to the field of detection of wind driven generators, in particular to a method for detecting the health state of a wind meter of a wind driven generator.

Background

One of the main sources of wind speed measurement data for a wind turbine is the wind meter of the wind turbine (wind generator) itself. The wind meter of the fan is influenced by factors such as working years, various natural environment factors, and wake flow generated by blades of other fans in a wind field, so that certain errors exist in the measurement precision, and the corresponding deviation exists between the generated energy calculated by the system by using the measured wind speed value and the actual generated energy undoubtedly. In the prior art, the fault of each fan anemoscope in the wind field is overhauled manually on time to reduce the error of the anemoscope when the wind speed is measured, however, the fault of each fan anemoscope is overhauled manually on time, which consumes more human resources and time resources, and has low overhauling efficiency.

Disclosure of Invention

In order to solve the problems of insufficient precision and low efficiency of manual measurement of a wind meter of a fan, the invention provides a method for detecting the health state of the wind meter of a wind driven generator, which has the following specific scheme:

a health state detection method for a wind meter of a wind driven generator comprises the following steps:

s1, obtaining SCADA data and filtering the data;

s2, preprocessing the filtered data;

s3, constructing a health scoring index and a judgment standard of the anemometer;

and S4, analyzing the health state of the anemometer according to the collected data.

Specifically, step S1 is to remove data that does not reflect the actual state of the wind turbine well, including data of the wind turbine generator in a shutdown state and a power-limiting operation state, and data of the wind speed lower than the first threshold and the generated power lower than the second threshold.

Specifically, step S2 specifically includes:

s2.1, acquiring the maximum wind speed V in the wind speed data_maxAnd minimum wind speed V_min；

S2.2, dividing the SCADA data into N intervals, wherein

E is a set speed; if the calculated N is a non-integer, taking the minimum positive integer larger than the calculated value;

s2.3, defining a numerical value K, wherein the initial value of the K is 1;

s2.4, calculating a wind speed range and a wind speed expected value in the Kth interval; calculating the expected power value of the generator in the Kth interval;

s2.5, determining that K is greater than N, if not, repeating step S2.4 after K is equal to K + 1; thus, the relation between the generator power and the wind speed is obtained.

Specifically, step S2.4 specifically includes:

s2.41, acquiring the wind speed range of the Kth wind speed interval,

[V_min+(K-1)V_S，V_min+KV_S]wherein

S2.42, calculating the expected wind speed value in the Kth interval:

in the formula v_iIs the wind speed; m is the number of different wind speed data in the Kth interval; p is a radical of_jIs the wind speed v_jThe probability of the value of (b) appearing in the Kth interval, whereinn is the number of data in the Kth interval, n_jThe wind speed in the Kth interval is v_jThe number of data of (2);

s2.43, calculating the expected value of the power of the generator in the Kth interval:

in the formula: p_iIs a power value; q is the number of different power data in the Kth interval; o_iIs a power P_iThe probability that the corresponding wind speed occurs in the kth interval,r is the number of power data in the K-th interval, r_iFor the power P in the Kth interval_iThe number of data of (2).

Specifically, step S3 specifically includes:

s3.1, constructing a relation function of anemometer measurement values and anemometer tower anemometer values;

s3.2, establishing a health scoring index of the anemometer according to the relation function in the S3.1;

and S3.3, constructing a judgment standard of the health state of the anemometer.

Specifically, step S3.1 specifically:

constructing a relation function of a anemometer measurement value X and a anemometer tower anemometer value Y,

SCADA data { x) for describing any set of inputs_i，y_iThe relationship between (i ═ 1, 2, … …, g); wherein a is_j(j-0, 1, 2, … …, z) are model coefficients,is y_iAn estimate of (d).

In particular, step S3.1 further comprises introducing a matrix

Wherein R is²Is the sum of the squares of the errors,

by substituting formula (6) into (5), A ═ X^TY(X^TX)^-1 (8)；

Calculating a from the above equation₀～a_zIs determined, input SCADA data { x_i，y_iThe relationship between them.

Specifically, step S3.2 specifically includes:

constructing a health scoring index S of the anemometer,

in the formula: a is_jModel coefficients calculated from the current SCADA data; b_jCalculating a model coefficient obtained from historical SCADA data of the fan in normal operation; x_maxAnd X_minThe maximum value and the minimum value in X are respectively, and K is the K-th interval.

Specifically, a criterion was constructed:

s is 0, the fan anemoscope runs normally;

s >0 indicates that the wind meter of the fan has faults, and the faults are more serious when the numerical value is larger.

Specifically, the method further includes step S3.4, specifically:

constructing fault classification according to corresponding deviation between the generated energy calculated by the wind speed value and the actual generated energy, and finally setting the generated energy deviation caused by the wind speed deviation caused by the fault as P_offsetActual power generation amount is P_cap；

When in useWhen S is 0, the fault grade is I, and the severity is normal;

when in useWhen S is 0, the fault grade is II, and the severity is slight;

when in useWhen S is 0, the fault grade is III, and the severity is severe;

when in useWhen S is equal to 0The failure rating is iv and the severity is very severe.

The invention has the beneficial effects that:

(1) the invention has low cost and great potential for monitoring the health condition of the wind meter of the fan, fully utilizes SCADA monitoring data, realizes the intellectualization, automation and real-time performance of detection, in addition, the invention provides a data preprocessing scheme to ensure the correlation between the wind speed of the wind meter and the wind speed of the wind measuring tower, so that the relationship between the wind speed and the wind measuring tower is more clearly shown, and meanwhile, the thought and the thought of the invention can be extended to more fan equipment components, thereby having great significance for the whole equipment detection of the fan.

(2) In the invention, an error matrix is introduced and reduced based on a least square method in the step S3.1 of data preprocessing, and the precision of a relation function is improved.

(3) According to the invention, the fault classification is established based on the relationship between the generated energy and the actual generated energy, so that the relationship between the health scoring index of the wind meter and the wind meter of the fan is more accurate.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.

FIG. 1 is a flow chart of the present invention;

FIG. 2 is a fault classification table of the present invention;

FIG. 3 is a plot of 1.5MW fan power after example data pre-processing;

FIG. 4 illustrates an example anemometer tower wind speed and a fan anemometer wind speed;

FIG. 5 is a comparison before and after a failure of a wind meter of a fan;

fig. 6 shows the scoring index S of the fan anemometer before and after the fan anemometer fails and the data fitting result.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

As shown in FIG. 1, the invention discloses a health state detection method for a wind meter of a wind driven generator, which comprises the following steps:

s1, obtaining SCADA data and filtering the data;

s2, preprocessing the filtered data;

s3, constructing a health scoring index and a judgment standard of the anemometer;

and S4, analyzing the health state of the anemometer according to the collected data.

Step S1 is specifically to remove data that cannot well reflect the real state of the wind turbine, including data of the wind turbine generator in a shutdown state and a power-limiting operation state, and data of the wind speed lower than the first threshold and the power generation power lower than the second threshold. Wherein the first threshold and the second threshold are set manually.

Step S2 specifically includes:

s2.1, acquiring the maximum wind speed V in the wind speed data_maxAnd minimum wind speed V_min；

S2.2, dividing the SCADA data into N intervals, wherein

E is a set speed; if the calculated N is a non-integer, taking the minimum positive integer larger than the calculated value;

s2.3, defining a numerical value K, wherein the initial value of the K is 1;

s2.4, calculating a wind speed range and a wind speed expected value in the Kth interval; calculating the expected power value of the generator in the Kth interval;

s2.5, determining that K is greater than N, if not, repeating step S2.4 after K is equal to K + 1; thus, the relation between the generator power and the wind speed is obtained.

Step S2.4 specifically includes:

s2.41, acquiring the wind speed range of the Kth wind speed interval,

[V_min+(K-1)V_S，V_min+KV_S]wherein

S2.42, calculating the expected wind speed value in the Kth interval:

in the formula v_iIs the wind speed; m is the number of different wind speed data in the Kth interval; p is a radical of_jIs the wind speed v_jThe probability of the value of (b) appearing in the Kth interval, whereinn is the number of data in the Kth interval, n_jThe wind speed in the Kth interval is v_jThe number of data of (2);

s2.43, calculating the expected value of the power of the generator in the Kth interval:

in the formula: p_iIs a power value; q is the number of different power data in the Kth interval; o_iIs a power P_iThe probability that the corresponding wind speed occurs in the kth interval,r is the number of power data in the K-th interval, r_iFor the power P in the Kth interval_iThe number of data of (2).

Step S3 specifically includes:

s3.1, constructing a relation function of anemometer measurement values and anemometer tower anemometer values;

s3.2, establishing a health scoring index of the anemometer according to the relation function in the S3.1;

and S3.3, constructing a judgment standard of the health state of the anemometer.

Step S3.1 is specifically:

constructing a relation function of a anemometer measurement value X and a anemometer tower anemometer value Y,

SCADA data { x) for describing any set of inputs_i，y_iThe relationship between (i ═ 1, 2, … …, g); wherein a is_j(j-0, 1, 2, … …, z) are model coefficients,is y_iAn estimate of (d).

Step S3.1 also includes introducing a matrix

Wherein R is²Is the sum of the squares of the errors,

by substituting formula (6) into (5), A ═ X^TY(X^TX)^-1(8)；

Calculating a from the above equation₀～a_zIs determined, input SCADA data { x_i，y_iThe relationship between them.

Step S3.2 specifically includes:

constructing a health scoring index S of the anemometer,

in the formula: a is_jModel coefficients calculated from the current SCADA data; b_jCalculating a model coefficient obtained from historical SCADA data of the fan in normal operation; x_maxAnd X_minThe maximum and minimum values in X, respectively.

Constructing a discrimination standard:

s is 0, the fan anemoscope runs normally;

s >0 indicates that the wind meter of the fan has faults, and the faults are more serious when the numerical value is larger.

Further comprising a step S3.4, specifically:

as shown in FIG. 2, a fault classification is constructed based on the corresponding deviation between the generated power calculated from the wind speed values and the actual generated power, and the deviation of the generated power finally caused by the deviation of the wind speed caused by the fault is P_offsetActual power generation amount is P_cap；

When in useWhen S is 0, the fault grade is I, and the severity is normal;

when in useWhen S is 0, the fault grade is II, and the severity is slight;

when in useWhen S is 0, the fault grade is III, and the severity is severe;

when in useThe failure grade is IV when S is 0, and the severity is very serious.

The method for testing the effect of the method comprises the following steps:

and comparing the SCADA data of the two 1.5MW fan anemometers in normal operation and after failure respectively, and calculating to obtain a score index S value.

Under normal conditions, the SCADA data of the wind meter of the fan and the recorded value of the wind meter of the wind tower have a relatively fixed relation, and once the relation between the SCADA data and the recorded value of the wind meter of the wind tower deviates, the abnormality of the wind meter is indicated. In the verification process, the data recorded by the other 1.5MW wind turbine generator cabin anemoscope before and after the fault and the change of the relation between the recorded values of the anemoscope tower are compared to verify the above assumptions. And calculating the fan health score index S.

As shown in fig. 4, the SCADA data before and after the failure of the wind meter of the wind turbine and the recorded wind speed value of the wind measuring tower are shown. From fig. 3, it is difficult to intuitively determine whether the relationship between the wind speed data of the wind turbine anemometer in normal or failure and the wind speed data of the wind measuring tower changes, so that the data needs to be processed. In the SCADA data, the maximum wind speed V measured before the fan fault occurs_maxThe minimum wind speed is V_min. According to the formula (1) and the requirement that N is an integer, determining that the value of N is 39, namely calculating the wind speed of the anemometer tower and the wind speed of the anemometer according to the formula (2) to obtain V_SAllocated to 39 intervals; then, expected values of the anemometer tower wind speed and the anemometer wind speed in 39 intervals were calculated, respectively. Data preprocessing, equation (5) a_jAnd j is 4, and the matrix A is obtained by solving the following steps: a when anemometer is not faulty₀～a₄Value, a when anemometer is faulty₀～a₄The value is obtained. The data were processed to yield fig. 5 and 6.

As shown in fig. 5 and 6, after data processing, information included in a scatter diagram which seems irregular before is extracted, and an operator can clearly judge the relationship between the wind speed of the wind meter of the wind turbine and the wind speed of the wind tower, wherein the wind speed of the wind meter of the wind turbine and the wind speed of the wind tower are in a positive correlation relationship. The relation between the wind speed of the wind measuring tower and the wind speed of the wind measuring instrument before and after a fault occurs is obviously changed, and S is equal to 0 before the fault occurs; after failure, S is 2.44. Therefore, the method provided by the invention can be used for diagnosing the fault of the wind meter of the fan.

Although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.