WIND TURBINE MONITORING DEVICE, METHOD, AND PROGRAM

• US 20120130678A1
• Filed: 08/31/2009
• Published: 05/24/2012
• Est. Priority Date: 08/31/2009
• Status: Active Grant

• Current Assignee: Mitsubishi Heavy Industries, Ltd.
• Original Assignee: Mitsubishi Heavy Industries, Ltd.
• Inventors: Ishioka, Masahito, Kobayashi, Yasuyuki
• Granted Patent: US 8,433,539 B2
• Time in Patent Office: Days
• US Class Current: 702/179
• CPC Class Codes: F03D 17/00 Monitoring or testing of wind motors, e.g. diagnostics testing during commissioning of wind motors F03D13/30
• CPC Class Codes: F05B 2270/80 Devices generating input signals, e.g. transducers, sensors, cameras or strain gauges
• CPC Class Codes: F05B 2270/803 Sampling thereof
• CPC Class Codes: G05B 23/024 Quantitative history assessment, e.g. mathematical relationships between available data; Functions therefor; Principal component analysis [PCA]; Partial least square [PLS]; Statistical classifiers, e.g. Bayesian networks, linear regression or correlation analysis; Neural networks

Abstract

Status monitoring of a wind turbine is automatically performed, and evaluation of that status is quantitatively performed according to an appropriate criterion. Provided is a wind turbine monitoring device including a first storage unit in which a diagnostic data file is stored; a second storage unit in which a normal data file is stored; a diagnostics setting unit that that extracts and sets a plurality of datasets to be used in diagnosis from the first storage unit and that extracts and sets a plurality of datasets to be used in diagnosis from the second storage unit; an index-value calculating unit that calculates status index values representing the status of the wind turbine using a statistical calculation method, on the basis of the set datasets in the diagnostic data file and the datasets in the reference data file; a fault determining unit that evaluates the status of the wind turbine on the basis of the status index values; and a notification unit that notifies the result of this evaluation.

Claims

1. A wind turbine monitoring device for monitoring the status of a wind turbine using attribute values created on the basis of measurement data measured by a plurality of sensors provided on the wind turbine, the wind turbine monitoring device comprising:

a first storage unit in which a plurality of attribute values that are associated with a measurement timestamp are stored by attribute type, and in which, when the attribute values associated with the same measurement timestamp serve as a single dataset, identifying information indicating a class category that is set according to attribute values of prescribed attributes is assigned to the dataset and stored;

a second storage unit in which a plurality of attribute values that are associated with a measurement timestamp are stored by attribute type, and in which, when the attribute values associated with the same measurement timestamp serve as a single dataset, identifying information that indicates a class category that is set according to the attribute values of prescribed attributes is assigned to the dataset, and the attribute values of the specific attributes forming the dataset belong to a prescribed standard range that is defined in advance;

a diagnostics setting unit that extracts and sets a plurality of the datasets to be used in diagnosis from the first storage unit and that extracts and sets a plurality of the datasets to be used in the diagnosis from the second storage unit;

an index-value calculating unit that calculates status index values representing the status of the wind turbine using a statistical calculation method, based on the datasets in a diagnostic data file and the datasets in a reference data file set by the diagnostics setting unit;

an evaluation unit that evaluates the status of the wind turbine on the basis of the status index values calculated by the index-value calculating unit; and

a notification unit that notifies a result of the evaluation by the evaluation unit.

2. A wind turbine monitoring device according to claim 1, wherein the attributes are broadly divided into three categories, which are an environment category related to the environment surrounding the wind turbine, a performance category related to the performance and power generating conditions of the wind turbine operation, and a characteristics category related to diagnosis of the operating status associated with various monitoring sites set on the wind turbine.

3. A wind turbine monitoring device according to claim 2, wherein the class categories are set according to the attribute values of prescribed attributes categorized into at least one of the environment category and the performance category.

4. A wind turbine monitoring device according to claim 1, wherein the plurality of datasets stored in the second storage unit are those in which only the datasets in which the attribute values associated with the specific attributes fall within the standard range set in advance are extracted from among the plurality of datasets created from the plurality of measurement data items collected from the wind turbine.

5. A wind turbine monitoring device according to claim 1, wherein the standard range is set according to the class category.

6. A wind turbine monitoring device according to claim 1, wherein the index-value calculating unit calculates the status index value by obtaining a characteristic distribution of the reference data and obtaining a characteristic distribution of the diagnostic data, which are set by the diagnostics setting unit, and by qualitatively obtaining the distance by which the characteristic distributions deviate from each other.

7. A wind turbine monitoring device according to claim 6, wherein the status index values calculated by the index-value calculating unit are Mahalanobis distances calculated using the Mahalanobis-Taguchi method.

8. A wind turbine monitoring device according to claim 1, wherein factorial analysis unit is provided that, when it is evaluated by the evaluation unit that a fault occurs, performs factorial analysis of that fault.

9. A wind-turbine-group monitoring system for monitoring the status of part or the entirety of a wind farm equipped with a plurality of wind turbines, comprising:

a wind turbine monitoring device according to claim 1, wherein the status of part or the entirety of the wind farm is monitored on the basis of a status index value of each of the wind turbines, obtained by the wind turbine monitoring device, and the operating performance of part or the entirety of the wind farm.

10. A wind turbine monitoring method which monitors the status of a wind turbine using attribute values created on the basis of measurement data measured by a plurality of sensors provided on the wind turbine, the wind turbine monitoring method comprising:

a step of creating a diagnostic data file in which a plurality of attribute values associated with a measurement timestamp are stored by attribute type;

an assigning step in which, when the attribute values associated with the same measurement timestamp in the diagnostic data file serve as a single dataset, the dataset is assigned identifying information indicating a class category that is set according to the attribute values of prescribed attributes;

a step of creating a reference data file in which attribute values related to specific attributes belong to a prescribed standard range that is defined in advance, and in which the attribute value of each attribute is associated with the measurement timestamp;

an assigning step in which, when the attribute values associated with the same measurement timestamp in the reference data file serve as a single dataset, the dataset is assigned identifying information indicating a class category that is set according to attribute values of prescribed attributes;

a step of extracting and setting a plurality of the datasets to be used in diagnosis from the diagnostic data file, and of extracting and setting a plurality of the datasets to be used in the diagnosis from the reference data file;

a step of calculating status index values representing the status of the wind turbine using a statistical calculation method, based on the set datasets in the diagnostic data file and the datasets in the reference data file;

a step of evaluating the status of the wind turbine on the basis of the status index values; and

a step of notifying the result of the evaluation.

11. A wind turbine monitoring program which monitors the status of a wind turbine using attribute values created on the basis of measurement data measured by a plurality of sensors provided on the wind turbine, the wind turbine monitoring program causing a computer to execute:

processing for creating a diagnostic data file in which a plurality of attribute values associated with a measurement timestamp are stored by attribute type;

assigning processing in which, when the attribute values associated with the same measurement timestamp in the diagnostic data file serve as a single dataset, the dataset is assigned identifying information indicating a class category that is set according to the attribute values of prescribed attributes;

processing for creating a reference data file in which attribute values related to specific attributes belong to a prescribed standard range that is defined in advance, and in which the attribute value of each attribute is associated with the measurement timestamp;

assigning processing in which, when the attribute values associated with the same measurement timestamp in the reference data file serve as a single dataset, the dataset is assigned identifying information indicating a class category that is set according to attribute values of prescribed attributes;

processing for extracting and setting a plurality of the datasets to be used in diagnosis from the diagnostic data file, and for extracting and setting a plurality of the datasets to be used in the diagnosis from the reference data file;

processing for calculating status index values representing the status of the wind turbine using a statistical calculation method, based on the set datasets in the diagnostic data file and the datasets in the reference data file;

processing for evaluating the status of the wind turbine on the basis of the status index values; and

processing for notifying the result of the evaluation.