CN114687930A - Wind turbine generator fault early warning closed-loop management and control system and method - Google Patents

Abstract

The invention discloses a wind turbine generator fault early warning closed-loop management and control system and a method, wherein the system comprises a data access layer, an early warning analysis layer, a fault diagnosis layer and a data interaction layer, SCADA data and CMS vibration data of a wind turbine generator are accessed into the early warning analysis layer, and when the wind turbine generator is in a fault early stage, an early warning model detects that the wind turbine generator is abnormal on a characteristic value representing a health state, and then the abnormality is caught, so that operation and maintenance personnel are guided to start intervention measures in time, and the further deterioration of the fault of the wind turbine generator is avoided; the method can realize early warning of faults, reduce unplanned shutdown of a wind field, realize deep excavation of fault hidden information, identify frequent faults to optimize a fault maintenance scheme, realize closed-loop management of the faults, form a case library and improve the life cycle health management of the wind turbine generator.

Description

A fault early warning closed-loop management and control system and method for a wind turbine

technical field

The invention relates to the technical field of on-line fault monitoring of wind turbines, in particular to a fault early warning closed-loop management and control system and method for wind turbines.

Background technique

At present, most of the faults given in the control system of wind turbines have already occurred, and no real fault warning has been realized; and no real closed-loop management has been implemented in fault handling. After the on-site defect processing is completed, there is no further Important information such as whether the deep digging fault is a recent frequent occurrence and its frequency, and whether other similar models in other wind farms have the same fault. In addition to SCADA data, wind turbines also have CMS vibration data, in which the application uses more of its amplitude changes, and does not identify its early and mid-term failures through data features. Most of the identified faults are reported from the control system. Such faults are after the fact, and there is no early warning of faults. The occurrence of serious faults will not only cause the shutdown of the unit, but also cause serious damage to the equipment; and when the fault occurs After the occurrence, it is generally tracked through various online platforms or excel forms. Once the fault processing is completed and closed, the data is not further excavated, and the value information deeply hidden in the database is not effectively extracted; In the dimension of analysis, it is more based on a certain wind farm base, and does not analyze the faults of all wind farms from a macro perspective, especially for wind turbines of the same type, the fault feedback is more valuable, due to the number of wind turbines. Many units will still have the same or similar faults after the faults are resolved, which indicates that the fault treatment is not complete or has not been completely cured. Such faults require the attention of on-site operation and maintenance personnel. The lack of corresponding technical means has led to a lack of attention. At the same time, it is difficult for operation and maintenance personnel to manually analyze the operation data of the past month or even the past 24 hours without borrowing advanced technical means. Identifying and catching exceptions in it is extremely inefficient.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a fault early warning closed-loop management and control system and method for wind turbines in order to solve the deficiencies in the prior art. The abnormality is detected, and then the abnormality is captured, so as to guide the operation and maintenance personnel to start the intervention measures in time to avoid the further deterioration of the wind turbine failure. At the same time, combined with CMS vibration data monitoring, the data transmitted from the wind farm can be fully utilized to realize multi-dimensional early warning of wind turbine failures.

In order to achieve the above purpose, the technical solution provided by the present invention is: a fault early warning closed-loop management and control system for a wind turbine, comprising:

The data access layer is used to connect the SCADA data and CMS vibration data of the wind turbines in the wind farm to the early warning analysis layer through the preset data transmission mechanism;

The early warning analysis layer preprocesses the connected SCADA data, selects the measurement points from the preprocessed SCADA data through an algorithm to model the early warning model, and inputs the operating data of the wind turbine into the early warning model to obtain the wind power The health state conclusion of the unit is output to the fault diagnosis layer; at the same time, the vibration signal characteristic quantity of the vibration signal is extracted after cleaning the connected CMS vibration data, and then the normal unit is marked by pre-analysis;

The fault diagnosis layer performs comprehensive diagnosis and analysis on the conclusions output by the early warning analysis layer, and outputs the diagnosis information to the data interaction layer;

The data interaction layer is used to import data to optimize the system and export and display early warning diagnostic reports.

Further, the data access layer specifically performs the following operations:

Enter the SCADA data files in CSV format in the wind farm into MySQL or Oracle relational database, convert the CMS vibration data in the wind farm into CSV format files or TXT format files through data conversion tools, and then enter them into MySQL or Oracle relational databases type database.

Further, the early warning analysis layer specifically performs the following operations:

The abnormal data or unusable data in the SCADA data are eliminated and the data that the unit is in normal operation is screened, and the pre-processed SCADA data is filtered through the screening points to model the early warning model, where the measurement points are the ones that can characterize the wind turbine. After the training of the early warning model is completed, the health degree of the wind turbine in the current time period can be calculated by reading other SCADA data, and a critical value of the health degree can be set;

Clean the CMS vibration data, generate new vibration signal data of each measuring point of the transmission chain through frequency resampling, or clean the CMS vibration data by stripping the invalid vibration signal data; The eigenvectors and frequency components of each component are marked out, and the vector and frequency component index parameters of each measuring point of the transmission chain are calculated. Label the normal units through pre-analysis.

Further, the fault diagnosis layer specifically performs the following operations:

By analyzing the list of all the measuring points that cause the health of the wind turbine to be less than the critical value of the health, the measuring points of the wind turbine with abnormal conditions and the names of the wind turbine equipment with abnormal conditions are displayed. At the same time, the vibration signal characteristics are analyzed, combined with SCADA process parameters and According to the conclusion output by the early warning analysis layer, the diagnostic information is finally output, and the diagnostic information includes the conclusion of the final diagnosis of the unit, the degree of unit failure, the maintenance suggestion of the unit, and the development trend of the failure.

Further, the early warning and diagnosis report includes a report on the failure of the wind turbine, the health status of the wind turbine, the statistics of the wind turbine prone and recurring failures, a report on major component failures, and an analysis curve for the deterioration trend of major components.

A wind turbine fault early warning closed-loop management and control method provided by the present invention uses the above-mentioned wind turbine fault early warning closed-loop management and control system, including the following steps:

S1. The wind turbine fault early warning closed-loop management and control system is connected to the SCADA data and CMS vibration data of the wind turbine;

S2, start the early warning analysis of the wind turbine fault early warning closed-loop management and control system according to the data accessed in step S1, calculate the health degree of the wind turbine by running the early warning model, and identify the abnormal vibration fault of the wind turbine according to the CMS vibration data;

S3. According to the identified fault of the wind turbine, determine the severity level of the fault, and check whether the fault has occurred on this wind turbine in the wind turbine fault early warning closed-loop management and control system, that is, whether it is a frequent fault, if it is a frequent fault If a fault occurs, the maintenance scheme for this frequent fault needs to be determined in combination with the previous maintenance scheme for the frequent fault; if it is not a frequent fault, the corresponding maintenance scheme is directly given for the faults of different severity levels;

S4. After the maintenance is completed, enter the relevant information of the fault in the wind turbine fault early warning closed-loop management and control system for recording.

Further, in step S4, the failure-related information includes a picture of the failure, the cause of the failure, and the spare parts and assembly dimensions replaced during the actual maintenance process.

Compared with the prior art, the present invention has the following advantages and beneficial effects:

The invention can realize early warning of faults, reduce unplanned shutdown of wind farms, and at the same time, can realize in-depth excavation of hidden fault information, identify frequent faults to optimize fault maintenance plans, realize fault closed-loop management, form a case database, and improve wind turbine generators. Life-cycle health management.

Description of drawings

Figure 1 is a framework diagram of a closed-loop management and control system for wind turbine fault early warning.

Detailed ways

The present invention will be further described below in conjunction with specific embodiments.

Referring to Figure 1, the wind turbine fault early warning closed-loop management and control system provided in this embodiment includes:

The data access layer is used to connect the SCADA data and CMS vibration data in the wind farm to the early warning analysis layer through the preset data transmission mechanism. Specifically, the following operations are performed:

Enter the SCADA data files in CSV format in the wind farm into MySQL or Oracle relational database, convert the CMS vibration data in the wind farm into CSV format files or TXT format files through data conversion tools, and then enter them into MySQL or Oracle relational databases type database.

The early warning analysis layer preprocesses the connected SCADA data, selects the measurement points from the preprocessed SCADA data through an algorithm to model the early warning model, and inputs the operating data of the wind turbine into the early warning model to obtain the wind power The health status conclusion of the unit is output to the fault diagnosis layer; at the same time, the vibration signal characteristic quantity of the vibration signal is extracted after cleaning the connected CMS vibration data, and then the normal unit is marked by pre-analysis. The early-warning analysis layer includes the fan early-warning analysis module And the CMS vibration data analysis module, which does the following:

The abnormal data or unusable data in the SCADA data are eliminated and the data that the unit is in normal operation is screened, and the pre-processed SCADA data is filtered through the screening points to model the early warning model, where the measurement points are the ones that can characterize the wind turbine. After the training of the early warning model is completed, the health degree of the wind turbine in the current time period can be calculated by reading other SCADA data, and a critical value of the health degree can be set;

Clean the CMS vibration data, generate new vibration signal data of each measuring point of the transmission chain through frequency resampling, or clean the CMS vibration data by stripping the invalid vibration signal data; The eigenvectors and frequency components of each component are marked out, and the vector and frequency component index parameters of each measuring point of the transmission chain are calculated. Label the normal units through pre-analysis.

The fault diagnosis layer performs comprehensive diagnosis and analysis on the conclusions output by the early warning analysis layer, and outputs the diagnosis information to the data interaction layer. The fault diagnosis layer includes a fan fault diagnosis module and a CMS diagnosis module, and specifically performs the following operations:

By analyzing the list of all the measuring points that cause the health of the wind turbine to be less than the critical value of the health, the measuring points of the wind turbine with abnormal conditions and the names of the wind turbine equipment with abnormal conditions are displayed. At the same time, the vibration signal characteristics are analyzed, combined with SCADA process parameters and According to the conclusion output by the early warning analysis layer, the diagnostic information is finally output, and the diagnostic information includes the conclusion of the final diagnosis of the unit, the degree of unit failure, the maintenance suggestion of the unit, and the development trend of the failure.

The data interaction layer is used to import data to optimize the system and export and display early warning diagnosis reports, which include wind turbine failure reports, wind turbine health status, wind turbine prone and recurring failure statistics, major component failure reports and Deterioration trend analysis curve of major components.

This embodiment discloses a wind turbine fault early warning closed-loop management and control method. Taking the generator bearing failure of wind turbine #009 in the wind farm as an example, the above-mentioned wind turbine fault early warning closed-loop management and control system is used, including the following steps:

S1. The wind turbine fault early warning closed-loop management and control system is connected to the SCADA data and CMS vibration data of the wind turbine #009;

S2, start the early warning analysis of the wind turbine fault early warning closed-loop management and control system according to the data accessed in step S1, calculate the health degree of wind turbine #009 by running the early warning model, and identify that the generator bearing temperature of wind turbine #009 is high, and Identify the abnormal vibration fault of wind turbine #009 according to the CMS vibration data, and identify the abnormal vibration amplitude and frequency spectrum of the generator bearing;

S3. Determine the severity of the fault according to the identified fault of wind turbine #009. If the fault is minor, wind turbine #009 can wait for a suitable window for maintenance; if the fault is serious, wind turbine #009 must be shut down for maintenance. , to avoid further deterioration of the fault; and check whether the fault has occurred on this wind turbine #009 in the wind turbine fault early warning closed-loop management and control system, that is, whether it is a frequent fault. The maintenance plan for frequent failures is used to determine this maintenance plan; if it is not a frequent failure, the corresponding maintenance plan is directly given for the failures of different severity levels;

S4. After the maintenance is completed, enter the fault-related information in the wind turbine fault early warning closed-loop management and control system for recording; the fault-related information includes the fault picture, the cause of the fault, and the spare parts and assembly dimensions replaced during the actual maintenance process. Through the on-site inspection, it was found that the cause of the abnormality of the generator bearing was the poor lubrication of the generator. After the lubricating oil was refilled on the site, the wind turbine #009 returned to normal operation.

In addition, the system is used to check whether such failures are frequent failures. If so, it is necessary to carry out a comprehensive census of the wind farm. At the same time, it is also possible to further analyze whether the failure occurs frequently only on this model. If so, then It is possible that the cause of this failure is a design defect of the model, which requires subsequent improvements to the model.

The above-mentioned embodiments are only preferred embodiments of the present invention, and are not intended to limit the scope of implementation of the present invention. Therefore, any changes made according to the shape and principle of the present invention should be included within the protection scope of the present invention.

Claims (7)

1. The utility model provides a wind turbine generator system trouble early warning closed loop management and control system which characterized in that includes:

the data access layer is used for accessing the SCADA data and the CMS vibration data of the wind turbine generator set in the wind field into the early warning analysis layer through a preset data transmission mechanism;

the early warning analysis layer is used for preprocessing the accessed SCADA data, screening out measuring points from the preprocessed SCADA data through an algorithm to perform modeling of an early warning model, inputting the operation data of the wind turbine generator into the early warning model to obtain a health state conclusion of the wind turbine generator, and outputting the health state conclusion to the fault diagnosis layer; meanwhile, the accessed CMS vibration data are cleaned, and then vibration signal characteristic quantities are extracted, and then normal units are marked in a pre-analysis mode;

the fault diagnosis layer is used for carrying out comprehensive diagnosis and analysis on the conclusion output by the early warning analysis layer and outputting diagnosis information to the data interaction layer;

and the data interaction layer is used for importing data to optimize the system and exporting and displaying the early warning diagnosis report.

2. The wind turbine generator system fault early warning closed-loop management and control system according to claim 1, wherein the data access layer specifically performs the following operations:

and recording the SCADA data file in the CSV format in the wind field into a MySQL or Oracle relational database, converting the CMS vibration data in the wind field into a CSV format file or a TXT format file through a data conversion tool, and recording the CSV format file or the TXT format file into the MySQL or Oracle relational database.

3. The wind turbine generator system fault early warning closed-loop management and control system according to claim 1, wherein the early warning analysis layer specifically performs the following operations:

rejecting abnormal data or unusable data in the SCADA data, screening data of the set in a normal operation state, carrying out early warning model modeling on the preprocessed SCADA data through screened measuring points, wherein the measuring points are combinations of parameters capable of representing the performance of the wind turbine generator, and after the early warning model training is finished, calculating the health degree of the wind turbine generator in the current time period by reading other SCADA data, and setting a health degree critical value;

carrying out data cleaning on CMS vibration data, generating new vibration signal data of each measuring point of a transmission chain by frequency resampling, or carrying out data cleaning on the CMS vibration data by stripping ineffective vibration signal data; and then marking characteristic vectors and frequency components of all parts in the vibration signal data by a frequency scanning algorithm, calculating index parameters of all measuring points and frequency components of the transmission chain, and marking the normal unit in a pre-analysis mode according to the characteristic vectors and the frequency components of all parts in the vibration signal data and the indexes of all measuring points and frequency components of the transmission chain.

4. The wind turbine generator system fault early warning closed-loop management and control system of claim 1, wherein the fault diagnosis layer specifically performs the following operations:

the method comprises the steps of displaying the measuring points with abnormal conditions and the names of the abnormal wind turbine equipment by analyzing a list of all measuring points which cause the health degree of the wind turbine to be smaller than a health degree critical value, analyzing vibration signal characteristics, combining SCADA process parameters and a conclusion output according to an early warning analysis layer, and finally outputting diagnosis information, wherein the diagnosis information comprises a conclusion of the final diagnosis of the wind turbine, a fault degree of the wind turbine, a maintenance suggestion of the wind turbine and a fault development trend.

5. The wind turbine generator system fault early warning closed-loop management and control system according to claim 1, wherein the early warning diagnosis report includes a wind turbine generator system fault report, a health state of the wind turbine generator system, statistics of faults which are easy to occur and are repeated, a major component fault report and a major component degradation trend analysis curve.

6. A wind turbine generator fault early warning closed-loop control method is characterized in that the wind turbine generator fault early warning closed-loop control system of any one of claims 1 to 5 is used, and comprises the following steps:

s1, accessing the wind turbine generator fault early warning closed-loop management and control system to SCADA data and CMS vibration data of the wind turbine generator;

s2, starting early warning analysis of the wind turbine fault early warning closed-loop management and control system according to the data accessed in the step S1, calculating the health degree of the wind turbine through an operation early warning model, and identifying the abnormal vibration fault of the wind turbine according to CMS vibration data;

s3, determining the severity level of the fault according to the identified fault of the wind turbine generator, checking whether the fault occurs on the wind turbine generator in a wind turbine generator fault early warning closed-loop management and control system, namely whether the fault belongs to frequent faults, and if the fault belongs to frequent faults, determining the maintenance scheme in combination with the previous maintenance scheme for maintaining the frequent faults; if the faults do not belong to frequent faults, corresponding maintenance schemes are directly provided for the faults with different severity levels;

and S4, after the maintenance is finished, recording the relevant information of the fault in the wind turbine generator fault early warning closed-loop management and control system for recording.

7. The wind turbine generator system fault early warning closed-loop management and control method according to claim 6, wherein in step S4, the fault related information includes a fault picture, a fault reason, spare parts and assembly size for replacement in an actual maintenance process.

Images