WO2021027294A1

WO2021027294A1 - Method and apparatus for improving wind power system data quality

Info

Publication number: WO2021027294A1
Application number: PCT/CN2020/082958
Authority: WO
Inventors: 杨晓茹; 鲍亭文; 王旻轩; 樊静; 金超
Original assignee: 北京天泽智云科技有限公司
Priority date: 2019-08-14
Filing date: 2020-04-02
Publication date: 2021-02-18
Also published as: WO2021027011A1; CN110489852A

Abstract

Disclosed in the present invention are a method and apparatus for improving wind power system data quality. The method comprises: determining error data in acquired system data and an error type thereof, the system data comprising data corresponding to each time point within a specific period of time, and the data of each time point comprising one or more data segments corresponding to different sensors; processing the error data in the system data according to the error type to obtain processed system data; and correcting the processed system data on the basis of a mechanism model to obtain corrected system data. By using the present invention, the quality of system data acquired by a wind power system can be effectively improved.

Description

Method and device for improving data quality of wind power system

Technical field

The invention relates to the field of data processing, in particular to a method and device for improving the data quality of a wind power system.

Background technique

The predictive maintenance system in the field of wind power has formed a full life cycle management system from terminal data collection, algorithm model building, and predictive maintenance. It provides early warning of wind power failure, stable operation of wind power, safe grid connection, and saving operation and maintenance costs. All provided technical guidance.

However, the operating conditions of wind turbines are complex and changeable, and the amount of data for the condition monitoring of the turbines is large. Due to factors such as turbine shutdown, load shedding, communication noise, and equipment failures, a large number of abnormal data will be generated, such as repeated values, Invalid data such as zero value, or collected data that does not conform to physical laws, etc., seriously affects the accuracy of the predictive maintenance model of wind power, leading to false alarms and omissions in the model early warning results. At the same time, due to the mismatch between the knowledge of experts in the field of wind power and the knowledge of algorithm modelers, the abnormal data screening during model modeling is incomplete, which increases the difficulty of early warning model modeling.

In summary, there are still many shortcomings in the collection, management, analysis, and mining of wind turbine operating data in the prior art. The quality difference of collected data cannot be accurately identified, and the correct selection and reasonable optimization of rough data cannot be effectively supported. Quality cannot be guaranteed. If these data are used directly without processing, the statistical characteristics of wind power will be distorted, which will affect the operation status of the wind turbine and the predictive maintenance results of the operation characteristics.

Summary of the invention

The embodiment of the present invention provides a method and device for improving the data quality of a wind power system, which can effectively process collected system data and improve the quality of source data.

To this end, the present invention provides the following technical solutions:

A method for improving data quality of a wind power system, the method comprising:

Determine the error data and error types in the collected system data; the system data includes: data corresponding to each time point in a certain time period, and the data at each time point includes one or more data segments corresponding to different sensors;

Processing the error data in the system data according to the error type to obtain processed system data;

Based on the mechanism model, the processed system data is corrected to obtain the corrected system data.

Optionally, the error types include any one or more of the following: missing errors, type errors, numerical errors, rule errors, and repeated errors;

The processing of the error data in the system according to the error type includes any one or more of the following:

Data filling is performed on the data segment whose error type is missing error;

Data type conversion or deletion of the data segment whose error type is type error;

Delete the data segment whose error type is numerical error;

Data conversion or deletion of data segments whose error types are rule errors;

For the data segment whose error type is repeated error, if all data segments corresponding to different sensors continue to repeat for more than the set time, the repeated data within the set time will be deleted; if the error type is repeated error data segment is the wind speed parameter , Or wind direction parameter, or temperature parameter, delete the repeated data segment.

Optionally, the performing data padding on the data segment whose error type is a missing error includes:

If the data segment whose error type is missing error is a wind speed parameter or a power parameter, the data shall be filled according to the wind power model;

Otherwise, the data is filled by interpolation.

Optionally, the correcting the processed system data based on the mechanism model includes:

Determine the abnormal wind speed parameter in the processed system data by using the environmental wind model, and delete the abnormal wind speed parameter;

Correcting the power parameters in the processed system data by using the wind power model;

The energy conservation model is used to determine the abnormal temperature parameter in the processed system data, and delete the abnormal temperature parameter.

Optionally, the using the wind power model to correct the power parameter in the processed system data includes:

Check in turn whether the data at each time point conforms to the wind power model;

If the data at the current time point does not conform to the wind power model, the power parameters in the current time point data are corrected according to the wind speed parameters and the wind turbine status code parameters in the current time point data.

Optionally, the correcting the power parameter in the data at the current time point according to the wind speed parameter and the wind turbine status code parameter in the data at the current time point includes:

If the wind speed parameter in the data at the current time point is less than the cut-in wind speed, modify the power parameter in the data at the current time point to 0;

If the wind speed parameter in the data at the current time point is greater than the cut-in wind speed and less than the rated wind speed, delete the power parameter in the data at the current time point;

If the wind speed parameter in the data at the current time point is greater than the rated wind speed, check whether the wind turbine status code in the data at the current time point is a limited power code;

If so, modify the power parameter in the data at the current time point to the limited power;

Otherwise, modify the power parameter in the data at the current time point to full power.

Optionally, the method further includes:

Determining abnormal data in the revised system data based on a mathematical model;

Remove the abnormal data to obtain optimized system data.

Optionally, the determining abnormal data in the corrected system data based on a mathematical model includes:

The abnormal data in the corrected system data is determined based on any one or more of the following mathematical models: a cluster model, a residual model, and a cluster model.

Optionally, the clustering model includes any one or more of the following: k-means model, DBSCAN model;

The input of the clustering model is the revised system data, and the output is the data category and quantity corresponding to different time points.

Optionally, the residual model includes any one or more of the following: linear regression, support vector machine, decision tree, neural network;

The input of the residual model is the corrected system data, and the output is the residuals corresponding to different time points.

Optionally, determining the abnormal data in the corrected system based on the cluster system includes:

Regarding the wind turbine system as a cluster model, determine whether the difference between each column of data of all wind turbines contained in each model in the cluster model and the corresponding average parameter of the wind farm exceeds the set threshold. If it exceeds the set threshold, determine the column of data The data at the time point is abnormal data.

A device for improving the data quality of a wind power system, the device comprising:

The error data detection module is used to determine the error data and the error type in the collected system data; the system data includes: data corresponding to each time point within a certain period of time, and each time point data includes one or more corresponding Data segments of different sensors;

A data processing module, configured to process the error data in the system data according to the error type to obtain processed system data;

The data correction module is used to correct the processed system data based on the mechanism model to obtain the corrected system data.

The data processing module is specifically configured to perform any one or more of the following processing on the error data in the system:

Delete the data segment whose error type is numerical error;

Data conversion or deletion of data segments whose error types are rule errors;

If all data segments corresponding to different sensors continue to repeat more than the set time, then delete the repeated data within the set time;

If the data segment whose error type is repeated error is wind speed parameter, or wind direction parameter, or temperature parameter, delete the repeated data segment.

Optionally, the data processing module performs data padding on data segments whose error types are missing errors in the following manner:

Otherwise, the data is filled by interpolation.

Optionally, the data correction module includes:

The wind speed parameter correction unit is used to determine the abnormal wind speed parameter in the processed system data by using the environmental wind model, and delete the abnormal wind speed parameter.

A power parameter correction unit, configured to use a wind power model to correct the power parameters in the processed system data;

The temperature parameter correction unit is used to determine the abnormal temperature parameter in the processed system data by using the energy conservation model, and delete the abnormal temperature parameter;

Optionally, the power parameter correction unit includes:

The check subunit is used to check whether the data at each time point conforms to the wind power model;

The correction subunit is used to correct the power parameters in the data at the current time point according to the wind speed parameters and the wind turbine status code parameters in the data at the current time point after the checking subunit determines that the data at the current time point does not conform to the wind power model .

Optionally, the correction subunit is specifically configured to modify the power parameter in the data at the current time point to 0 when the wind speed parameter in the data at the current time point is less than the cut-in wind speed; When the wind speed parameter is greater than the cut-in wind speed and less than the rated wind speed, delete the power parameter in the data at the current time point; when the wind speed parameter in the data at the current time point is greater than the rated wind speed, check whether the fan status code in the data at the current time point is It is a limited power code; if it is, the power parameter in the data at the current time point is modified to the limited power; otherwise, the power parameter in the data at the current time point is modified to the full power.

Optionally, the device further includes:

An abnormal data detection module for determining abnormal data in the corrected system data based on a mathematical model;

The abnormal data cleaning module is used to remove the abnormal data to obtain optimized system data.

Optionally, the abnormal data detection module includes any one or more of the following modules:

The first detection module is configured to detect abnormal data in the corrected system data based on the clustering model;

The second detection module is configured to detect abnormal data in the corrected system data based on the residual model;

The third detection module is used to detect abnormal data in the corrected system data based on the cluster model.

The input of the clustering model is the corrected system data, and the output is the data category and its quantity corresponding to different time points.

Optionally, the third detection module is specifically configured to use the wind turbine system as a cluster model to determine whether the difference between each column of data of all wind turbines included in each model in the cluster model and the corresponding average parameter of the wind field exceeds the set value. Set the threshold. If it exceeds the set threshold, the data at the time point of the column of data is determined to be abnormal data.

An electronic device including: one or more processors and memories;

The memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the aforementioned method.

A readable storage medium having instructions stored thereon, and the instructions are executed to implement the aforementioned method.

The method and device for improving the data quality of the wind power system provided by the embodiments of the present invention determine the error data and the error type in the collected wind power system data, and first process the error data according to the different error types to obtain the processed data System data; Then, according to the processed system data, the data is corrected based on the mechanism model to obtain the corrected system data. The solution of the embodiment of the present invention not only processes the erroneous data caused in the process of data collection and transmission, but also corrects the system data based on the mechanism model according to the characteristics of the wind power system data, so that the finally obtained system data has higher quality , Effectively avoiding low-quality source data from interfering with subsequent system operation and maintenance management, and ensuring the normal operation of wind turbines.

Further, abnormal data detection can be performed on the corrected system data based on one or more mathematical models, and the detected abnormal data can be removed, so that the system data can be further optimized, and the quality of the system data can be better guaranteed.

Description of the drawings

In order to more clearly describe the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed in the embodiments. Obviously, the drawings in the following description are only those described in the present invention. For some of the embodiments, for those of ordinary skill in the art, other drawings may be obtained based on these drawings.

Fig. 1 is a flowchart of a method for improving data quality of a wind power system according to an embodiment of the present invention;

FIG. 2 is a flowchart of using a wind power model to correct power parameters in system data in an embodiment of the present invention;

FIG. 3 is another flowchart of a method for improving data quality of a wind power system according to an embodiment of the present invention;

Fig. 4 is a structural block diagram of the device for improving the data quality of the wind power system implemented by the present invention;

FIG. 5 is a structural block diagram of a data correction module in an embodiment of the present invention;

Fig. 6 is another structural block diagram of a device for improving data quality of a wind power system according to an embodiment of the present invention.

detailed description

In order to enable those skilled in the art to better understand the solutions of the embodiments of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings and implementation manners.

The embodiment of the present invention provides a method and device for improving the data quality of a wind power system. According to the collected wind power system data, the error data and the error type are determined, and the error data is first processed according to different error types. The system data; then for the processed system data, the data is corrected based on the mechanism model to obtain the corrected system data.

As shown in Figure 1, it is a flowchart of a method for improving data quality of a wind power system according to an embodiment of the present invention, which includes the following steps:

Step 101: Determine the error data and the error type in the collected system data.

The system data includes: data corresponding to each time point in a certain time period, and the data at each time point includes one or more data segments corresponding to different sensors.

Due to the large number of wind power system units and different operating states, the amount of data collected by monitoring is very large; in addition, the data transmission process will inevitably be affected by factors such as noise interference, causing errors in some data in the collected system data. Bad data. Due to the different causes and objects of the error data, the error type of different error data will be different.

In practical applications, the existing technology can be used to determine the error data and the error type in the system data. Among them, the error types are mainly as follows:

Missing error, that is, the data in a certain data segment is empty;

Type error, that is, the data type of a data segment does not match the actual type of the data segment. For example, the value of a data segment should be 12.5, but it is recorded as the string ‘12.5’;

Numerical error, that is, the value of a certain data segment is not within the specified range. For example, the ambient temperature value is recorded as 100°C, which obviously does not conform to common sense;

Rule error, that is, a certain data segment does not meet the set rules, such as the fan status code should be an integer between 0 and n, and the fan start and stop status should be 0 or 1;

Duplicate error, that is, a certain data repeats within a certain period of time.

Of course, according to actual application requirements, there may be other error types, which are not limited in the embodiment of the present invention.

In addition, in order to more accurately and comprehensively determine the error data therein, the embodiment of the present invention also provides a method for determining the error data and its error type. This method is not only performed horizontally for the system data collected within a certain period of time. Detection is to detect the data at each time point, and also to perform longitudinal detection, that is, to detect each data segment within a certain time period. The method specifically includes: horizontal detection and vertical detection; among them:

Horizontal detection refers to the detection of data at each time point in sequence with time points as a unit, marking the position of abnormal data segments in the data, and recording error types;

Longitudinal detection refers to sequentially detecting each data segment within a certain period of time using data segments as a unit, and recording abnormal data segments and error types.

Wherein, the detection of data at each time point in the horizontal detection mainly includes:

Detect whether there is a missing data segment in the data; if so, mark the position of the missing data segment, and record the error type as: missing error;

Detect whether the type of each data segment in the data is correct; if not, mark the position of the data segment, and record the error type as: type error;

Detect whether the value of each data segment in the data is within the specified range; if not, mark the position of the data segment, and record the type of error as: numerical error;

It is detected whether each data segment in the data meets the set rule; if it does not, the position of the data segment is marked, and the error type is recorded as: rule error.

In the longitudinal detection, detecting each data segment within the certain time period includes:

Detect the repeatability of the data segment in the certain time period, mark the repeated data segment, and record the error type as: repeated error;

Detect the continuity of the data segment within the certain time period, mark the missing data segment, and record the error type as: missing error.

Step 102: Process the error data in the system data according to the error type to obtain processed system data.

Different types of error data can be handled in different ways. In the embodiment of the present invention, the following processing methods can be adopted for the several types of errors mentioned above:

(1) Data filling is performed on the data segment whose error type is missing error, specifically in the following situations:

If a certain row of data (that is, data corresponding to a certain point in time) or a certain column of data (that is, corresponding to a certain data segment within a certain period of time) in the system data are all missing, then the row or column of data is not Fill, delete the row or column directly.

If an individual data segment in a row or a column is missing, different filling measures need to be taken according to the system parameters represented by the missing data segment. For example, if the error type is the missing data segment is the wind speed parameter or power parameter, Then the data can be filled according to the wind power model, that is to say, the value of the missing data segment is calculated according to the other data segment and the wind power model corresponding to the time point of the missing data segment, and the value is filled to the corresponding position; The wind power model is a functional relationship between wind speed and active power of a wind turbine, and is related to the factory design parameters of the wind turbine itself.

If the error type is missing and the wrong data segment is not a wind speed parameter or a power parameter, since wind power system data usually uses a high sampling rate for data collection, data can be filled by interpolation, and the time point of the missing data segment can be used. The corresponding data segments before and after are interpolated and filled. If the corresponding data segments before and after the time point to which the missing data segment belongs are also missing, random numbers can be used for interpolation and filling, where the random numbers come from the 3sigma range of the normal distribution of the data segment.

(2) Data type conversion or deletion is performed on the data segment whose error type is type error.

(3) Delete the data segment whose error type is numerical error.

(4) Data conversion or deletion is performed on the data segment whose error type is rule error.

For example, the fan status code column should be an integer. If the corresponding data segment is not an integer, then its value will be rounded. If the corresponding data segment is a character string and the rounding process cannot be performed, the data segment will be deleted.

(5) For the data segment whose error type is repeated error: check whether all data segments corresponding to different sensors continue to repeat for more than a set time (for example, 30 minutes), if so, delete the repeated data within the set time; Otherwise, check whether the data segment whose error type is repeated error is a wind speed parameter, or a wind direction parameter, or a temperature parameter, and if it is, delete the repeated data segment.

It should be noted that in practical applications, any one or more of the above-mentioned processing methods can be adopted as required, and the processing sequence of different types of error data is not required. Of course, the embodiments of the present invention are not limited to the above-mentioned processing, and there may be corresponding processing methods for other error types.

Step 103: Correct the processed system data based on the mechanism model to obtain corrected system data.

According to the characteristics of the data collected by the wind power system, such as wind speed, active power, gearbox oil temperature, gearbox bearing temperature, generator bearing temperature, engine room temperature, engine room control cabinet temperature, etc., these data should theoretically conform to a certain relationship and satisfy Corresponding mechanism models, such as: wind power model, energy conservation model, environmental wind model, etc. Therefore, in the method of the embodiment of the present invention, the above-mentioned mechanism models can be used to modify some data segments in the system data, and there are mainly the following three processing methods:

(1) Using the environmental wind model to determine the abnormal wind speed parameter in the processed system data, and delete the abnormal wind speed parameter.

For example, if there is no change in wind speed, but the change in wind direction exceeds 1 degree, the corresponding wind direction parameter is abnormal data and it is deleted.

(2) Using the wind power model to correct the power parameters in the processed system data.

The specific process of using the wind power model to correct the power parameters in the processed system data can be seen in Figure 2, which includes the following steps:

Step 201: Check in turn whether the data at each time point conforms to the wind power model; if not, perform step 202; otherwise, end.

Step 202: Determine whether the wind speed parameter in the data at the current time point is less than the cut-in wind speed; if yes, go to step 203; otherwise, go to step 204;

Step 203: Modify the power parameter in the data at the current time point to 0;

Step 204: Determine whether the wind speed parameter in the data at the current time point is less than the rated wind speed; if yes, go to step 205; otherwise, go to step 206;

Step 205: Delete the power parameter in the data at the current time point;

Step 206, check whether the wind turbine status code in the data at the current time point is a limited power code; if yes, go to step 207; otherwise, go to step 208;

Step 207: Modify the power parameter in the data at the current time point to a limited power;

Step 208: Modify the power parameter in the data at the current time point to full power.

(3) Use the energy conservation model to determine the abnormal temperature parameter in the processed system data, and delete the abnormal temperature parameter.

For example, an energy conservation model is established for the approximately enclosed space such as the nacelle, control cabinet, and gear box in the wind turbine system. When the actual data collected does not conform to the energy conservation model, the corresponding data is deleted.

It should be noted that, in practical applications, the above-mentioned sequence of correcting system data using different models is not limited.

The method for improving the data quality of the wind power system provided by the embodiment of the present invention determines the error data and the error type in the collected wind power system data. First, the error data is processed according to the different error types to obtain the processed system data ; Then for the processed system data, data correction is made based on the mechanism model to obtain the corrected system data. The solution of the embodiment of the present invention not only processes the erroneous data caused in the process of data collection and transmission, but also corrects the system data based on the mechanism model according to the characteristics of the wind power system data, so that the finally obtained system data has higher quality , Effectively avoiding low-quality source data from interfering with subsequent system operation and maintenance management, and ensuring the normal operation of wind turbines.

As shown in Figure 3, it is another flow chart of the method for improving the data quality of the wind power system according to the embodiment of the present invention, which includes the following steps:

Step 301: Determine the error data and the error type in the collected system data.

Step 302: Process the error data in the system data according to the error type to obtain processed system data.

Step 303: Correct the processed system data based on the mechanism model to obtain corrected system data.

The above steps 301 to 303 are the same as the steps 101 to 103 in FIG. 1 and will not be repeated here.

Step 304: Determine abnormal data in the corrected system data based on a mathematical model.

Specifically, the abnormal data in the revised system data may be determined based on any one or more of the following mathematical models: a clustering model, a residual model, and a clustering model.

It should be noted that if the abnormal data is determined based on the above two or more mathematical models, the union of the abnormal data obtained according to different models can be used as the final abnormal data, or the abnormal data obtained from different models can be merged Calculation (such as weighted calculation), according to the calculation result to determine the final abnormal data.

The clustering model may specifically include any one or more of the following: k-means model, DBSCAN (Density-Based Spatial Clustering of Applications with Noise) model; the input of the clustering model is the corrected system data , The output is the data category and quantity corresponding to different time points.

If the number of a certain category is less than the set value (for example, 3), the category data is determined to be abnormal data.

The residual model may specifically include any one or more of the following: linear regression, support vector machine, decision tree, neural network; the input of the residual model is the corrected system data, and the output is different time points The corresponding residual.

If the residual error corresponding to a certain time point exceeds the set threshold, all data at that time point is determined to be abnormal data.

In a wind power plant, a wind farm often includes dozens of wind turbines, and the external environment of the wind turbines is similar. Therefore, the operating states of different wind turbines in the same wind farm are similar. Based on this feature, in the embodiment of the present invention, the wind turbine system is taken as a cluster model, and it is judged that each column of data of all wind turbines contained in each model in the cluster model and the corresponding average parameters of the wind field (such as temperature, wind speed, rotation speed, power) Etc.) Whether the difference value exceeds the set threshold, if it exceeds the set threshold, it is determined that the data at the time point of the column of data is abnormal data.

Step 305: Remove the abnormal data to obtain optimized system data.

The method for improving the data quality of the wind power system provided by the embodiment of the present invention not only processes the erroneous data caused in the data collection and transmission process of the wind power system, but also corrects the system data based on the mechanism model according to the characteristics of the wind power system data. One or more mathematical models perform abnormal data detection on the corrected system data, and remove the detected abnormal data, so that the system data is further optimized, and the quality of the system data can be better guaranteed.

Correspondingly, the embodiment of the present invention also provides a device for improving the data quality of the wind power system. As shown in FIG. 4, it is a structural block diagram of the device for improving the data quality of the wind power system implemented by the present invention.

In this embodiment, the device includes the following modules:

The error data detection module 401 is used to determine the error data and the error type in the collected system data; the system data includes: data corresponding to each time point in a certain time period, and the data at each time point includes one or more Corresponding to the data segment of different sensors;

The data processing module 402 is configured to process the error data in the system data according to the error type to obtain processed system data;

The data correction module 403 is used to correct the processed system data based on the mechanism model to obtain the corrected system data.

In practical applications, the above-mentioned error data detection module 401 can use existing technology to determine the error data and the error type in the system data. Among them, the error types mainly include the following types: missing errors, type errors, numerical errors, rule errors, and repeated errors. Of course, according to actual application requirements, there may be other error types, which are not limited in the embodiment of the present invention.

The meaning of the above error types has been explained in detail above, so I won't repeat them here.

In addition, in order to more accurately and comprehensively determine the error data therein, the above-mentioned error data detection module 401 can also determine the error data and its error type in the following manner:

Horizontal detection: using time point as a unit, sequentially detect data at each time point, mark the position of the abnormal data segment in the data, and record the error type;

Longitudinal detection: Take the data segment as a unit, sequentially detect each data segment within the certain time period, and record the abnormal data segment and the error type.

The above-mentioned horizontal detection and vertical detection methods and processes have been described in detail above, and will not be repeated here.

For the error data of different error types, the data processing module 402 can adopt different processing methods. For example, any one or more of the following processing can be performed on the error data in the system:

Data filling is performed on the data segment whose error type is missing error. For example, if the data segment with the error type of missing error is a wind speed parameter or power parameter, data filling is performed according to the wind power model; otherwise, data filling is performed through interpolation;

Data type conversion or deletion of the data segment whose error type is numerical error;

Delete the data segment whose error type is rule error;

If the data segment whose error type is repeated error is a wind speed parameter, or wind direction parameter, or temperature parameter, the repeated data segment is deleted.

According to the characteristics of the data collected by the wind power system, the data correction module 403 can use various mechanism models to correct some data segments in the system data. For example, a specific structure of the data correction module 403 is shown in Figure 5. Show, including the following units:

The wind speed parameter correction unit 433 is configured to determine the abnormal wind speed parameter in the processed system data by using the environmental wind model, and delete the abnormal wind speed parameter;

The power parameter correction unit 431 is configured to correct the power parameters in the processed system data by using the wind power model;

The temperature parameter correction unit 432 is configured to determine an abnormal temperature parameter in the processed system data by using an energy conservation model, and delete the abnormal temperature parameter.

The environmental wind model, wherein the power parameter correction unit includes: an inspection unit and a correction subunit, wherein:

The checking subunit is used to check whether the data at each time point conforms to the wind power model in turn;

The correction subunit is used for correcting the power in the data at the current time point according to the wind speed parameters and the wind turbine status code parameters in the data at the current time point after the checking subunit determines that the data at the current time point does not conform to the wind power model parameter. Specifically, when the wind speed parameter in the data at the current time point is less than the cut-in wind speed, the power parameter in the data at the current time point is modified to 0; when the wind speed parameter in the data at the current time point is greater than the cut-in wind speed and less than the rated wind speed , Delete the power parameter in the data at the current time point; when the wind speed parameter in the data at the current time point is greater than the rated wind speed, check whether the status code of the wind turbine in the data at the current time point is a limited power code; if so, change The power parameter in the data at the current time point is modified to the limited power; otherwise, the power parameter in the data at the current time point is modified to the full power. The specific correction process can refer to the embodiment shown in Figure 2 above.

The device for improving the data quality of the wind power system provided by the embodiment of the present invention determines the error data and the error type in the collected wind power system data, and first processes the error data according to the different error types to obtain the processed system data ; Then for the processed system data, data correction is made based on the mechanism model to obtain the corrected system data. The solution of the embodiment of the present invention not only processes the erroneous data caused in the process of data collection and transmission, but also corrects the system data based on the mechanism model according to the characteristics of the wind power system data, so that the finally obtained system data has higher quality , Effectively avoiding low-quality source data from interfering with subsequent system operation and maintenance management, and ensuring the normal operation of wind turbines.

As shown in FIG. 6, it is another structural block diagram of an apparatus for improving data quality of a wind power system according to an embodiment of the present invention.

Different from the embodiment shown in FIG. 4, in this embodiment, the device further includes the following modules:

An abnormal data detection module 404, configured to determine abnormal data in the corrected system data based on a mathematical model;

The abnormal data cleaning module 405 is used to remove the abnormal data to obtain optimized system data.

In practical applications, the abnormal data detection module 404 may determine the abnormal data in the corrected system data based on one or more mathematical models. For example, a specific structure of the abnormal data detection module 404 may include But not limited to any one or more of the following modules:

The clustering model may include, but is not limited to, any one or more of the following: k-means model, DBSCAN model; the input of the clustering model is the corrected system data, and the output is corresponding to different time points The data category and its quantity.

The residual model may include, but is not limited to, any one or more of the following: linear regression, support vector machine, decision tree, neural network; the input of the residual model is the corrected system data, and the output is different The residual error corresponding to the time point.

The third detection module is specifically used to use the wind turbine system as a cluster model to determine whether the difference between the data of all the wind turbines contained in each model in the cluster model and the corresponding average parameter of the wind farm exceeds a set threshold, and if it exceeds Set the threshold to determine that the data at the time point of the column of data is abnormal data.

It should be noted that, for each embodiment of the device for improving the data quality of the wind power system, since the function implementation of each module and unit is similar to that of the corresponding method, the description of each embodiment of the dialog generating device is relatively simple. For related details, please refer to the description of the corresponding part of the method embodiment.

The device for improving the data quality of the wind power system provided by the embodiment of the present invention not only processes the erroneous data caused in the data collection and transmission process of the wind power system, but also corrects the system data based on the mechanism model according to the characteristics of the wind power system data. One or more mathematical models perform abnormal data detection on the corrected system data, and remove the detected abnormal data, so that the system data is further optimized, and the quality of the system data can be better guaranteed.

It should be noted that the terms "first" and "second" in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and not necessarily used to describe a specific sequence or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations of them are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to the clearly listed Those steps or units may include other steps or units that are not clearly listed or are inherent to these processes, methods, products, or equipment.

The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. Moreover, the system embodiments described above are only illustrative, and the modules and units described as separate components may or may not be physically separated, that is, they may be located on one network unit, or may be distributed to multiple On the network unit. Some or all of the modules may be selected according to actual needs to achieve the objectives of the solutions of the embodiments. Those of ordinary skill in the art can understand and implement it without creative work.

A person of ordinary skill in the art can understand that all or part of the steps in the above-mentioned method embodiments can be implemented by a program instructing relevant hardware. The program can be stored in a computer-readable storage medium, which is referred to as storage herein. Medium, such as: ROM/RAM, magnetic disk, optical disk, etc.

Correspondingly, an embodiment of the present invention also provides a device for improving the data quality of a wind power system. The device is an electronic device, such as a mobile terminal, a computer, a tablet device, a medical device, a fitness device, or a personal computer. Digital assistants, etc. The electronic device may include one or more processors and memories; wherein the memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to implement the foregoing method.

The embodiments of the present invention are described in detail above, and specific implementations are used to illustrate the present invention. The descriptions of the above embodiments are only used to help understand the methods and devices of the present invention, and they are only part of the embodiments of the present invention. Not all examples. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative work shall fall within the protection scope of the present invention, and the contents of this specification should not be construed as limiting the present invention. Therefore, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

A method for improving data quality of a wind power system, characterized in that the method includes:

Determine the error data and error types in the collected system data; the system data includes: data corresponding to each time point in a certain time period, and the data at each time point includes one or more data segments corresponding to different sensors;

Processing the error data in the system data according to the error type to obtain processed system data;

Based on the mechanism model, the processed system data is corrected to obtain the corrected system data.
The method according to claim 1, wherein the error types include any one or more of the following: missing errors, type errors, numerical errors, rule errors, and repeated errors;

The processing of the error data in the system according to the error type includes any one or more of the following:

Data filling is performed on the data segment whose error type is missing error;

Data type conversion or deletion of the data segment whose error type is type error;

Delete the data segment whose error type is numerical error;

Data conversion or deletion of data segments whose error types are rule errors;

For the data segment whose error type is repeated error, if all data segments corresponding to different sensors continue to repeat for more than the set time, the repeated data within the set time will be deleted; if the error type is repeated error data segment is the wind speed parameter , Or wind direction parameter, or temperature parameter, delete the repeated data segment.
The method according to claim 2, wherein said performing data padding on the data segment whose error type is missing error comprises:

If the data segment whose error type is missing error is a wind speed parameter or a power parameter, the data shall be filled according to the wind power model;

Otherwise, the data is filled by interpolation.
The method according to claim 1, wherein the correcting the processed system data based on the mechanism model comprises:

Determine the abnormal wind speed parameter in the processed system data by using the environmental wind model, and delete the abnormal wind speed parameter;

Correcting the power parameters in the processed system data by using the wind power model;

The energy conservation model is used to determine the abnormal temperature parameter in the processed system data, and delete the abnormal temperature parameter.
The method according to claim 4, wherein said correcting the power parameter in the processed system data by using a wind power model comprises:

Check in turn whether the data at each time point conforms to the wind power model;

If the data at the current time point does not conform to the wind power model, the power parameters in the current time point data are corrected according to the wind speed parameters and the wind turbine status code parameters in the current time point data.
The method according to claim 5, wherein the correcting the power parameter in the data at the current time point according to the wind speed parameter and the wind turbine status code parameter in the data at the current time point comprises:

If the wind speed parameter in the data at the current time point is less than the cut-in wind speed, modify the power parameter in the data at the current time point to 0;

If the wind speed parameter in the data at the current time point is greater than the cut-in wind speed and less than the rated wind speed, delete the power parameter in the data at the current time point;

If the wind speed parameter in the data at the current time point is greater than the rated wind speed, check whether the wind turbine status code in the data at the current time point is a limited power code;

If so, modify the power parameter in the data at the current time point to the limited power;

Otherwise, modify the power parameter in the data at the current time point to full power.
The method according to any one of claims 1 to 6, wherein the method further comprises:

Determining abnormal data in the revised system data based on a mathematical model;

Remove the abnormal data to obtain optimized system data.
The method according to claim 7, wherein the determining abnormal data in the corrected system data based on a mathematical model comprises:

The abnormal data in the revised system data is determined based on any one or more of the following mathematical models: clustering model, residual model, and clustering model.
The method according to claim 8, wherein the determining the abnormal data in the corrected system based on the cluster system comprises:

Regarding the wind turbine system as a cluster model, determine whether the difference between each column of data of all wind turbines contained in each model in the cluster model and the corresponding average parameter of the wind farm exceeds the set threshold. If it exceeds the set threshold, determine the column of data The data at the time point is abnormal data.
A device for improving data quality of a wind power system, characterized in that the device comprises:

The error data detection module is used to determine the error data and the error type in the collected system data; the system data includes: data corresponding to each time point within a certain period of time, and each time point data includes one or more corresponding Data segments of different sensors;

A data processing module, configured to process the error data in the system data according to the error type to obtain processed system data;

The data correction module is used to correct the processed system data based on the mechanism model to obtain the corrected system data.
The device according to claim 10, wherein the error type includes any one or more of the following: missing error, type error, numerical error, rule error, and repeated error;

The data processing module is specifically configured to perform any one or more of the following processing on the error data in the system:

Data filling is performed on the data segment whose error type is missing error;

Data type conversion or deletion of the data segment whose error type is type error;

Delete the data segment whose error type is numerical error;

Data conversion or deletion of data segments whose error types are rule errors;

If all data segments corresponding to different sensors continue to repeat more than the set time, then delete the repeated data within the set time;

If the data segment whose error type is repeated error is a wind speed parameter, or wind direction parameter, or temperature parameter, the repeated data segment is deleted.
The device according to claim 11, wherein the data processing module performs data padding on data segments whose error types are missing errors in the following manner:

If the data segment whose error type is missing error is a wind speed parameter or a power parameter, the data shall be filled according to the wind power model;

Otherwise, the data is filled by interpolation.
The device according to claim 10, wherein the data correction module comprises:

The wind speed parameter correction unit is used to determine the abnormal wind speed parameter in the processed system data by using the environmental wind model, and delete the abnormal wind speed parameter.

A power parameter correction unit, configured to use a wind power model to correct the power parameters in the processed system data;

The temperature parameter correction unit is used to determine the abnormal temperature parameter in the processed system data by using the energy conservation model, and delete the abnormal temperature parameter;
The device according to claim 13, wherein the power parameter correction unit comprises:

The check subunit is used to check whether the data at each time point conforms to the wind power model;

The correction subunit is used to correct the power parameters in the data at the current time point according to the wind speed parameters and the wind turbine status code parameters in the data at the current time point after the checking subunit determines that the data at the current time point does not conform to the wind power model .
The device of claim 14, wherein:

The correction subunit is specifically used to modify the power parameter in the data at the current time point to 0 when the wind speed parameter in the data at the current time point is less than the cut-in wind speed; the wind speed parameter in the data at the current time point is greater than the cut-in wind speed When the wind speed is less than the rated wind speed, delete the power parameter in the data at the current time point; when the wind speed parameter in the data at the current time point is greater than the rated wind speed, check whether the fan status code in the data at the current time point is a limited power code ; If yes, modify the power parameter in the data at the current time point to the limited power; otherwise, modify the power parameter in the data at the current time point to full power.
The device according to any one of claims 10 to 15, wherein the device further comprises:

An abnormal data detection module for determining abnormal data in the corrected system data based on a mathematical model;

The abnormal data cleaning module is used to remove the abnormal data to obtain optimized system data.
The device according to claim 16, wherein the abnormal data detection module comprises any one or more of the following modules:

The first detection module is configured to detect abnormal data in the corrected system data based on the clustering model;

The second detection module is configured to detect abnormal data in the corrected system data based on the residual model;

The third detection module is used to detect abnormal data in the corrected system data based on the cluster model.
The device according to claim 17, wherein:

The third detection module is specifically used to use the wind turbine system as a cluster model to determine whether the difference between each column of data of all wind turbines included in each model in the cluster model and the corresponding average parameter of the wind field exceeds a set threshold, If the threshold is exceeded, it is determined that the data at the time point of the column of data is abnormal data.