[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

WO2016086360A1 - Wind farm condition monitoring method and system - Google Patents

Wind farm condition monitoring method and system Download PDF

Info

Publication number
WO2016086360A1
WO2016086360A1 PCT/CN2014/092813 CN2014092813W WO2016086360A1 WO 2016086360 A1 WO2016086360 A1 WO 2016086360A1 CN 2014092813 W CN2014092813 W CN 2014092813W WO 2016086360 A1 WO2016086360 A1 WO 2016086360A1
Authority
WO
WIPO (PCT)
Prior art keywords
historical
real time
data
wind turbine
status
Prior art date
Application number
PCT/CN2014/092813
Other languages
French (fr)
Inventor
Niya CHEN
Rongrong Yu
Yao Chen
Original Assignee
Abb Technology Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Abb Technology Ltd filed Critical Abb Technology Ltd
Priority to PCT/CN2014/092813 priority Critical patent/WO2016086360A1/en
Publication of WO2016086360A1 publication Critical patent/WO2016086360A1/en

Links

Images

Classifications

    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03D—WIND MOTORS
    • F03D17/00—Monitoring or testing of wind motors, e.g. diagnostics
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03D—WIND MOTORS
    • F03D7/00—Controlling wind motors 
    • F03D7/02—Controlling wind motors  the wind motors having rotation axis substantially parallel to the air flow entering the rotor
    • F03D7/04—Automatic control; Regulation
    • F03D7/042—Automatic control; Regulation by means of an electrical or electronic controller
    • F03D7/043—Automatic control; Regulation by means of an electrical or electronic controller characterised by the type of control logic
    • F03D7/045—Automatic control; Regulation by means of an electrical or electronic controller characterised by the type of control logic with model-based controls
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F03—MACHINES OR ENGINES FOR LIQUIDS; WIND, SPRING, OR WEIGHT MOTORS; PRODUCING MECHANICAL POWER OR A REACTIVE PROPULSIVE THRUST, NOT OTHERWISE PROVIDED FOR
    • F03D—WIND MOTORS
    • F03D7/00—Controlling wind motors 
    • F03D7/02—Controlling wind motors  the wind motors having rotation axis substantially parallel to the air flow entering the rotor
    • F03D7/04—Automatic control; Regulation
    • F03D7/042—Automatic control; Regulation by means of an electrical or electronic controller
    • F03D7/048—Automatic control; Regulation by means of an electrical or electronic controller controlling wind farms
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B23/00—Testing or monitoring of control systems or parts thereof
    • G05B23/02—Electric testing or monitoring
    • G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
    • G05B23/0208—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the configuration of the monitoring system
    • G05B23/021—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterized by the configuration of the monitoring system adopting a different treatment of each operating region or a different mode of the monitored system, e.g. transient modes; different operating configurations of monitored system
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B23/00—Testing or monitoring of control systems or parts thereof
    • G05B23/02—Electric testing or monitoring
    • G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
    • G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
    • G05B23/0243—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model
    • G05B23/0254—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults model based detection method, e.g. first-principles knowledge model based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F05—INDEXING SCHEMES RELATING TO ENGINES OR PUMPS IN VARIOUS SUBCLASSES OF CLASSES F01-F04
    • F05B—INDEXING SCHEME RELATING TO WIND, SPRING, WEIGHT, INERTIA OR LIKE MOTORS, TO MACHINES OR ENGINES FOR LIQUIDS COVERED BY SUBCLASSES F03B, F03D AND F03G
    • F05B2260/00—Function
    • F05B2260/84—Modelling or simulation
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02E—REDUCTION OF GREENHOUSE GAS [GHG] EMISSIONS, RELATED TO ENERGY GENERATION, TRANSMISSION OR DISTRIBUTION
    • Y02E10/00—Energy generation through renewable energy sources
    • Y02E10/70—Wind energy
    • Y02E10/72—Wind turbines with rotation axis in wind direction

Definitions

  • the present application relates to a wind farm condition monitoring method and system.
  • Wind power has become the most popular renewable energy promising to replace traditional pollutive thermal power generation because of its rich resources, mature technology, and zero emission.
  • Global wind turbine installation had reached 318GW at the end of 2013.
  • O&M operation and maintenance
  • electricity sale loss develop to be more and more pressing issues.
  • one aspect of the present invention provides a wind farm condition monitoring method, comprising the following steps:
  • historical data acquiring step acquiring historical SCADA data, and/or historical sensor data received from sensors which are installed in a wind farm, and historical wind turbine status which indicate wind turbines historical status in the wind farm, wherein the historical SCADA data covers operation data of the wind farm;
  • model training step selecting algorithm for all types of historical data sources according to pre-defined basic rules, wherein the historical data sources include the historical SCADA data and/or the historical sensor data, training different models for different historical data sources using the algorithm selected for different types of historical data sources to establish relationship between the historical data sources and the historical wind turbine status;
  • model diagnosing step acquiring real time SCADA data or real time sensor data as real time data source, selecting the trained model depending on type of the real time data source, inputting the real time data source to the selected trained model, obtaining the real time wind turbine status, which estimate wind turbines real time status in the wind farm, from the selected trained model.
  • the historical data acquiring step further comprises:
  • verifying the wind turbine status to identify which data section of the historical SCADA data and/or the historical sensor data is in normal status and which data section of the historical SCADA data or the historical sensor data is in defective status.
  • the model training step comprises:
  • selecting the most effective trained model from the models having same type of historical data source selecting the primary algorithm corresponding to the most effective trained model as the algorithm selected for the type of historical data source, retraining models for each historical data source using the algorithm selected for the type of historical data source.
  • the model diagnosing step comprises:
  • model updating step updating the trained models with the real time data source and the real time wind turbine status.
  • the model updating step comprises:
  • the wind turbine status covers:
  • health alarming step activating an alarm device when the health condition of the real time wind turbine status is diagnosed as defective status.
  • Another aspect of the present invention provides a computer program comprising computer program code adapted to perform all of the steps of any one of the above when run on a computer.
  • a further aspect of the present invention provides a computer program according to the above, embodied on a computer-readable medium.
  • historical data acquiring module used for acquiring historical SCADA data, and/or historical sensor data received from sensors which are installed in a wind farm, and historical wind turbine status which indicate wind turbines historical status in the wind farm, wherein the historical SCADA data covers operation data of the wind farm;
  • model training module used for selecting algorithm for all types of historical data sources according to pre-defined basic rules, wherein the historical data sources include the historical SCADA data and/or the historical sensor data, training different models for different historical data sources using the algorithm selected for different types of historical data sources to establish relationship between the historical data sources and the historical wind turbine status;
  • model diagnosing module used for acquiring real time SCADA data and/or real time sensor data as real time data source, selecting the trained model depending on type of the real time data source, inputting the real time data source to the selected trained model, obtaining the real time wind turbine status, which estimate wind turbines real time status in the wind farm, from the selected trained model.
  • the historical data acquiring module further comprises:
  • module used for verifying the wind turbine status to identify which data section of the historical SCADA data and/or the historical sensor data is in normal status and which data section of the historical SCADA data or the historical sensor data is in defective status.
  • the model training module comprises:
  • module used for training models for each historical data source using the primary algorithm to establish relationship between the historical data sources and the historical wind turbine status, wherein the inputs of each model are the historical data sources, and the outputs of each model are the historical wind turbine status;
  • module used for selecting the most effective trained model from the models having same type of historical data source, selecting the primary algorithm corresponding to the most effective trained model as the algorithm selected for the type of historical data source, retraining models for each historical data source using the algorithm selected for the type of historical data source.
  • the model diagnosing module comprises:
  • module used for acquiring real time SCADA data and/or real time sensor data as real time data source
  • module used for selecting the trained model depending on type of the real time data source
  • model updating module updating the trained models with the real time data source and the real time wind turbine status.
  • model updating module comprises:
  • module used for retraining the trained models with the real time data source and the real time wind turbine report wherein the inputs of each model are the real time data sources, the outputs of each model are the real time wind turbine status, and the algorithm used is algorithm selected for the type of real time data source.
  • the wind turbine status covers:
  • health alarming module used for activating an alarm device when the health condition of the real time wind turbine status is diagnosed as defective status.
  • the present invention trained the model with the historical SCADA data and/or historical sensor data as data sources, and historical wind turbine status as output, by using different algorithm for different type of data source.
  • the present invention proposed a self-adaptive CMS which is featured in terms of flexibility in using any available data source and automatic adjustment according to wind farm condition.
  • the methods designed in this invention are quite generic which can be applied to different applications (e.g. different wind farm configurations) .
  • the present invention is compatible with current CMS product, no matter it is sensor system or SCADA system, without hardware modification, and even has potential ability for other possible data source and future techniques.
  • Fig. 1 shows a flow-process diagram illustrating a wind farm condition monitoring method in accordance with the present invention
  • Fig. 2 shows a flow-process diagram illustrating the preferred embodiment of the present invention
  • Fig. 3 shows a structural module drawing of a wind farm condition monitoring system.
  • Fig. 1 shows a flow-process diagram illustrating a wind farm condition monitoring method, comprising the following steps:
  • step 101 acquiring historical SCADA data, and/or historical sensor data received from sensors which are installed in a wind farm, and historical wind turbine status which indicate wind turbines historical status in the wind farm, wherein the historical SCADA data covers operation data of the wind farm;
  • step 102 selecting algorithm for all types of historical data sources according to pre-defined basic rules, wherein the historical data sources include the historical SCADA data and/or the historical sensor data, training different models for different historical data sources using the algorithm selected for different types of historical data sources to establish relationship between the historical data sources and the historical wind turbine status;
  • step 103 acquiring real time SCADA data and/or real time sensor data as real time data source, selecting the trained model depending on type of the real time data source, inputting the real time data source to the selected trained model, obtaining the real time wind turbine status, which estimate wind turbines real time status in the wind farm, from the selected trained model.
  • the SCADA data is collected from the Supervisory Control And Data Acquisition (SCADA) system.
  • SCADA Supervisory Control And Data Acquisition
  • the SCADA system is a power automation monitoring system. It performs data collection, monitoring control and process control of the power system.
  • the data source includes SCADA data and sensor data.
  • the sensor data includes: vibration sensor data, acoustic sensor data, etc.
  • the types of data sources include: type of SCADA data, type of vibration sensor data, type of acoustic sensor data, etc.
  • selecting algorithm for all types of historical data sources includes: selecting algorithm for type of SCADA data, selecting algorithm for type of vibration sensor data, selecting algorithm for type of acoustic sensor data, etc.
  • the present invention is a kind of self-adaptive condition monitoring method and system for wind farm that is flexible in data sources and can automatically adjust to different wind farm/turbine condition.
  • step 101 further comprises:
  • verifying the wind turbine status to identify which data section of the historical SCADA data and/or the historical sensor data is in normal status and which data section of the historical SCADA data or the historical sensor data is in defective status.
  • the embodiment divides the historical SCADA data or historical sensor data into normal status section and defective status section to facilitate the models making a correct wind turbine report.
  • step 102 comprises:
  • selecting the most effective trained model from the models having same type of historical data source selecting the primary algorithm corresponding to the most effective trained model as the algorithm selected for the type of historical data source, retraining models for each historical data source using the algorithm selected for the type of historical data source.
  • the primary algorithm can be different. For example: for vibration sensor data, the primary algorithm is more often chose from time series methods, and for SCADA data, the primary algorithm can be selected from data mining methods.
  • the embodiment verifies the effectiveness (e.g. detection rate) of the trained models and find out the most effective trained model from the models having same type of historical data source. For example, the embodiment selects the most effective trained model from the models with SCADA data as data source, or selects most effective trained model from the models with vibration sensor data as data source.
  • the effectiveness e.g. detection rate
  • step 103 comprises:
  • model updating step updating the trained models with the real time data source and the real time wind turbine status.
  • the models are trained with the historical data source and historical wind turbine status.
  • the historical data source will have some different with the real time data source, especially when the trained model run a long time.
  • the embodiment add model updating step in order to amend the trained model.
  • model updating step run after the step 103.
  • the model updating step comprises:
  • Updating the trained model means retraining the trained models with the real time data source and the real time wind turbine report. And the algorithm used for retrain is the same as before.
  • the wind turbine status covers:
  • health alarming step activating an alarm device when the health condition of the real time wind turbine status is diagnosed as defective status.
  • Fig. 2 shows a flow-process diagram illustrating the preferred embodiment of the present invention.
  • Left side block represents the flexibility in data sources and the algorithm/model selection process; while right side block represents model updating process with newly collected online data (real time data source) . More specifically, the whole concept can be implemented by the following steps:
  • Step 201 Collecting all available monitoring data (from SCADA system, already installed vibration/acoustic/oil sensor, etc. ) and fault information from real wind farm operation database as wind turbine report:
  • Step 202 Selecting proper algorithm for available data source, according to pre-defined basic rules and test results of trained models on collected historical data:
  • ⁇ Training models based on primary selected algorithms to be applicable for wind farm diagnosis and prognosis.
  • the inputs of each model are historical data, and the output is the health condition of target wind turbine, which component is defective (if wind turbine is defective) and the fault information like location, type and severity.
  • Step 203 Updatina models trained bv selected algorithm based on newly collected online data consistently:
  • Step 204 Using the updated models to implement wind turhine diagnosis/proanosis:
  • this self-adaptive CMS solution can provide wind farm a flexible, complete and accurate condition monitoring service without any hardware modification.
  • on-line refers to online real time monitor.
  • off-line refers to offline models training.
  • Fig. 3 shows a structural module drawing of a wind farm condition monitoring system, comprising the following modules:
  • historical data acquiring module 301 used for acquiring historical SCADA data, and/or historical sensor data received from sensors which are installed in a wind farm, and historical wind turbine status which indicate wind turbines historical status in the wind farm, wherein the historical SCADA data covers operation data of the wind farm;
  • model training module 302 used for selecting algorithm for all types of historical data sources according to pre-defined basic rules, wherein the historical data sources include the historical SCADA data and/or the historical sensor data, training different models for different historical data sources using the algorithm selected for different types of historical data sources to establish relationship between the historical data sources and the historical wind turbine status;
  • model diagnosing module 303 used for acquiring real time SCADA data and/or real time sensor data as real time data source, selecting the trained model depending on type of the real time data source, inputting the real time data source to the selected trained model, obtaining the real time wind turbine status, which estimate wind turbines real time status in the wind farm, from the selected trained model.

Landscapes

  • Engineering & Computer Science (AREA)
  • Mechanical Engineering (AREA)
  • Physics & Mathematics (AREA)
  • Sustainable Energy (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Sustainable Development (AREA)
  • General Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Mathematical Physics (AREA)
  • Wind Motors (AREA)

Abstract

A wind farm condition monitoring method and system, comprises: acquiring historical SCADA data, and/or historical sensor data received from sensors which are installed in a wind farm, and historical wind turbine status which indicate wind turbines historical status (101); selecting algorithm for all types of historical data sources according to pre-defined basic rules, training different models for different historical data sources to establish relationship between the historical data sources and the historical wind turbine status (102); acquiring real time SCADA data and/or real time sensor data as real time data source, selecting the trained model depending on type of the real time data source, inputting the real time data source to the selected trained model, obtaining the real time wind turbine status (103).

Description

WIND FARM CONDITION MONITORING METHOD AND SYSTEM TECHNICAL FIELD
The present application relates to a wind farm condition monitoring method and system.
BACKGROUND ART
Wind power has become the most popular renewable energy promising to replace traditional pollutive thermal power generation because of its rich resources， mature technology， and zero emission. Global wind turbine installation had reached 318GW at the end of 2013. With rapid installation increase of wind farms， expensive O&M (operation and maintenance) cost and downtime electricity sale loss develop to be more and more pressing issues. Taking a 2MW wind turbine as example， as evaluated by master thesis of KTH， about 248.4kUSD annual cost arises， incl. 242.7kUSD O&M cost and 5.7kUSD electricity sale loss.
Under such environment， it’s desired by the market that a kind of CMS product is able to real-time detect the defects of wind turbine， analyze the fault type， and position the defective part， before it evolves to real failure. Now the available products in market with such functionality are all based on specified sensors， e.g. vibration， acoustic， etc. However， the situation is， firstly there is already a SCADA system， which monitors many processing parameters， is installed in each turbine， thus would be huge waste if not used in CMS； secondly， normally different kinds of sensors/SCADA system are installed in different wind farms， therefore none current CMS product is generally compatible without hardware modification； thirdly， an advanced CMS should have potential space for other possible data source and future techniques.
SUMMARY
Accordingly， one aspect of the present invention provides a wind farm condition monitoring method， comprising the following steps：
historical data acquiring step： acquiring historical SCADA data， and/or historical sensor data received from sensors which are installed in a wind farm， and  historical wind turbine status which indicate wind turbines historical status in the wind farm， wherein the historical SCADA data covers operation data of the wind farm；
model training step： selecting algorithm for all types of historical data sources according to pre-defined basic rules， wherein the historical data sources include the historical SCADA data and/or the historical sensor data， training different models for different historical data sources using the algorithm selected for different types of historical data sources to establish relationship between the historical data sources and the historical wind turbine status；
model diagnosing step： acquiring real time SCADA data or real time sensor data as real time data source， selecting the trained model depending on type of the real time data source， inputting the real time data source to the selected trained model， obtaining the real time wind turbine status， which estimate wind turbines real time status in the wind farm， from the selected trained model.
Preferably， the historical data acquiring step further comprises：
verifying the wind turbine status to identify which data section of the historical SCADA data and/or the historical sensor data is in normal status and which data section of the historical SCADA data or the historical sensor data is in defective status.
Preferably， the model training step comprises：
building up the algorithm base and basic rules for algorithm primary screening；
selecting a primary algorithm from the algorithm base for each historical data source according to the basic rules；
training models for each historical data source using the primary algorithm to establish relationship between the historical data sources and the historical wind turbine status， wherein the inputs of each model are the historical data sources， and the outputs of each model are the historical wind turbine status；
verifying the effectiveness of the trained models using the historical data source and the historical wind turbine report；
selecting the most effective trained model from the models having same type of historical data source， selecting the primary algorithm corresponding to the most effective trained model as the algorithm selected for the type of historical data  source， retraining models for each historical data source using the algorithm selected for the type of historical data source.
Conveniently， the model diagnosing step comprises：
acquiring real time SCADA data and/or real time sensor data as real time data source；
selecting the trained model depending on type of the real time data source；
inputting the real time data source to the trained model；
running the trained model to implement the algorithm；
obtaining the real time wind turbine report from the trained model.
Preferably， further comprises：
model updating step： updating the trained models with the real time data source and the real time wind turbine status.
Conveniently， the model updating step comprises：
running the trained models for a predetermined period；
retraining the trained models with the real time data source and the real time wind turbine report， wherein the inputs of each model are the real time data sources， the outputs of each model are the real time wind turbine status， and the algorithm used is algorithm selected for the type of real time data source.
Conveniently， the wind turbine status covers：
health condition of wind turbine in the wind farm diagnosed as normal or defective status， and defective component with corresponding fault reason if health condition of wind turbine is diagnosed as defective status.
Conveniently， further comprises：
health alarming step： activating an alarm device when the health condition of the real time wind turbine status is diagnosed as defective status.
Another aspect of the present invention provides a computer program comprising computer program code adapted to perform all of the steps of any one of the above when run on a computer.
A further aspect of the present invention provides a computer program  according to the above， embodied on a computer-readable medium.
Another aspect of the present invention provides a wind farm condition monitoring system， comprising the following modules：
historical data acquiring module： used for acquiring historical SCADA data， and/or historical sensor data received from sensors which are installed in a wind farm， and historical wind turbine status which indicate wind turbines historical status in the wind farm， wherein the historical SCADA data covers operation data of the wind farm；
model training module： used for selecting algorithm for all types of historical data sources according to pre-defined basic rules， wherein the historical data sources include the historical SCADA data and/or the historical sensor data， training different models for different historical data sources using the algorithm selected for different types of historical data sources to establish relationship between the historical data sources and the historical wind turbine status；
model diagnosing module： used for acquiring real time SCADA data and/or real time sensor data as real time data source， selecting the trained model depending on type of the real time data source， inputting the real time data source to the selected trained model， obtaining the real time wind turbine status， which estimate wind turbines real time status in the wind farm， from the selected trained model.
Preferably， the historical data acquiring module further comprises：
module used for verifying the wind turbine status to identify which data section of the historical SCADA data and/or the historical sensor data is in normal status and which data section of the historical SCADA data or the historical sensor data is in defective status.
Preferably， the model training module comprises：
module used for building up the algorithm base and basic rules for algorithm primary screening；
module used for selecting a primary algorithm from the algorithm base for each historical data source according to the basic rules；
module used for training models for each historical data source using the  primary algorithm to establish relationship between the historical data sources and the historical wind turbine status， wherein the inputs of each model are the historical data sources， and the outputs of each model are the historical wind turbine status；
module used for verifying the effectiveness of the trained models using the historical data source and the historical wind turbine report；
module used for selecting the most effective trained model from the models having same type of historical data source， selecting the primary algorithm corresponding to the most effective trained model as the algorithm selected for the type of historical data source， retraining models for each historical data source using the algorithm selected for the type of historical data source.
Conveniently， the model diagnosing module comprises：
module used for acquiring real time SCADA data and/or real time sensor data as real time data source；
module used for selecting the trained model depending on type of the real time data source；
module used for inputting the real time data source to the trained model；
module used for running the trained model to implement the algorithm；
module used for obtaining the real time wind turbine report from the trained model.
Preferably， further comprises：
model updating module： updating the trained models with the real time data source and the real time wind turbine status.
Conveniently， the model updating module comprises：
module used for running the trained models for a predetermined period；
module used for retraining the trained models with the real time data source and the real time wind turbine report， wherein the inputs of each model are the real time data sources， the outputs of each model are the real time wind turbine status， and the algorithm used is algorithm selected for the type of real time data source.
Conveniently， the wind turbine status covers：
health condition of wind turbine in the wind farm diagnosed as normal or  defective status， and defective component with corresponding fault reason if health condition of wind turbine is diagnosed as defective status.
Conveniently， further comprises：
health alarming module： used for activating an alarm device when the health condition of the real time wind turbine status is diagnosed as defective status.
The present invention trained the model with the historical SCADA data and/or historical sensor data as data sources， and historical wind turbine status as output， by using different algorithm for different type of data source. As receiving the historical SCADA data and historical sensor data as data sources， the present invention proposed a self-adaptive CMS which is featured in terms of flexibility in using any available data source and automatic adjustment according to wind farm condition. The methods designed in this invention are quite generic which can be applied to different applications (e.g. different wind farm configurations) . And as more data source is accepted， including the sensor data from current CMS， the present invention is compatible with current CMS product， no matter it is sensor system or SCADA system， without hardware modification， and even has potential ability for other possible data source and future techniques.
BRIEF DESCRIPTION OF THE DRAWINGS
Fig. 1 shows a flow-process diagram illustrating a wind farm condition monitoring method in accordance with the present invention；
Fig. 2 shows a flow-process diagram illustrating the preferred embodiment of the present invention；
Fig. 3 shows a structural module drawing of a wind farm condition monitoring system.
DETAILED DESCRIPTION OF THE EMBODIMENTS
Hereinafter， the present invention is further introduced in detail by the particular embodiments in combination with the figures.
Fig. 1 shows a flow-process diagram illustrating a wind farm condition  monitoring method， comprising the following steps：
step 101： acquiring historical SCADA data， and/or historical sensor data received from sensors which are installed in a wind farm， and historical wind turbine status which indicate wind turbines historical status in the wind farm， wherein the historical SCADA data covers operation data of the wind farm；
step 102： selecting algorithm for all types of historical data sources according to pre-defined basic rules， wherein the historical data sources include the historical SCADA data and/or the historical sensor data， training different models for different historical data sources using the algorithm selected for different types of historical data sources to establish relationship between the historical data sources and the historical wind turbine status；
step 103： acquiring real time SCADA data and/or real time sensor data as real time data source， selecting the trained model depending on type of the real time data source， inputting the real time data source to the selected trained model， obtaining the real time wind turbine status， which estimate wind turbines real time status in the wind farm， from the selected trained model.
The SCADA data is collected from the Supervisory Control And Data Acquisition (SCADA) system. The SCADA system is a power automation monitoring system. It performs data collection， monitoring control and process control of the power system.
The data source includes SCADA data and sensor data. The sensor data includes： vibration sensor data， acoustic sensor data， etc. So the types of data sources include： type of SCADA data， type of vibration sensor data， type of acoustic sensor data， etc. In the step 102， selecting algorithm for all types of historical data sources includes： selecting algorithm for type of SCADA data， selecting algorithm for type of vibration sensor data， selecting algorithm for type of acoustic sensor data， etc.
As accepting variety of type of data sources and selecting algorithm based on the type of data sources， the present invention is a kind of self-adaptive condition monitoring method and system for wind farm that is flexible in data sources and can automatically adjust to different wind farm/turbine condition.
In one embodiment， step 101 further comprises：
verifying the wind turbine status to identify which data section of the historical SCADA data and/or the historical sensor data is in normal status and which data section of the historical SCADA data or the historical sensor data is in defective status.
The embodiment divides the historical SCADA data or historical sensor data into normal status section and defective status section to facilitate the models making a correct wind turbine report.
In one embodiment， step 102 comprises：
building up the algorithm base and basic rules for algorithm primary screening；
selecting a primary algorithm from the algorithm base for each historical data source according to the basic rules；
training models for each historical data source using the primary algorithm to establish relationship between the historical data sources and the historical wind turbine status， wherein the inputs of each model are the historical data sources， and the outputs of each model are the historical wind turbine status；
verifying the effectiveness of the trained models using the historical data source and the historical wind turbine report；
selecting the most effective trained model from the models having same type of historical data source， selecting the primary algorithm corresponding to the most effective trained model as the algorithm selected for the type of historical data source， retraining models for each historical data source using the algorithm selected for the type of historical data source.
There are many algorithms for model training， e.g. Artificial Neural Network， Support Vector Machine， Gaussian Processes， etc. The definition of basic rules should consider not only feature of algorithms but also the characteristics of data sources. For different available data sources， the primary algorithm can be different. For example： for vibration sensor data， the primary algorithm is more often chose from time series methods， and for SCADA data， the primary algorithm can be selected from data mining methods.
For the same type of data source， there may be more than one primary algorithm selected for training. So after training the models using the primary  algorithm， the embodiment verifies the effectiveness (e.g. detection rate) of the trained models and find out the most effective trained model from the models having same type of historical data source. For example， the embodiment selects the most effective trained model from the models with SCADA data as data source， or selects most effective trained model from the models with vibration sensor data as data source.
In one embodiment， step 103 comprises：
acquiring real time SCADA data and/or real time sensor data as real time data source；
selecting the trained model depending on type of the real time data source；
inputting the real time data source to the trained model；
running the trained model to implement the algorithm；
obtaining the real time wind turbine report from the trained model.
In one embodiment， further comprises：
model updating step： updating the trained models with the real time data source and the real time wind turbine status.
The models are trained with the historical data source and historical wind turbine status. The historical data source will have some different with the real time data source， especially when the trained model run a long time. The embodiment add model updating step in order to amend the trained model.
It should be noted that the model updating step run after the step 103.
In one embodiment， the model updating step comprises：
running the trained models for a predetermined period；
retraining the trained models with the real time data source and the real time wind turbine report， wherein the inputs of each model are the real time data sources， the outputs of each model are the real time wind turbine status， and the algorithm used is algorithm selected for the type of real time data source.
Updating the trained model means retraining the trained models with the real time data source and the real time wind turbine report. And the algorithm used for retrain is the same as before.
In one embodiment， the wind turbine status covers：
health condition of wind turbine in the wind farm diagnosed as normal or defective status， and defective component with corresponding fault reason if health condition of wind turbine is diagnosed as defective status.
In one embodiment， further comprises：
health alarming step： activating an alarm device when the health condition of the real time wind turbine status is diagnosed as defective status.
Fig. 2 shows a flow-process diagram illustrating the preferred embodiment of the present invention.
In this figure， the method is divided into two parts： offline preparation work and online adjusting part. Left side block represents the flexibility in data sources and the algorithm/model selection process； while right side block represents model updating process with newly collected online data (real time data source) . More specifically， the whole concept can be implemented by the following steps：
Step 201： Collecting all available monitoring data (from SCADA system，  already installed vibration/acoustic/oil sensor， etc. ) and fault information from real  wind farm operation database as wind turbine report：
·SCADA data covering operation data of typical wind turbines in typical wind farms
·Data from additional sensors installed like vibration sensor that can reflect more details of wind farm condition
·Verified wind turbine status corresponding to the monitoring data above specifying which data section is normal status and which is defective status
·Organize the data and fault information to proper format compatible for following data process
·Note that this step should be done off-line instead of on-line.
Step 202： Selecting proper algorithm for available data source， according to  pre-defined basic rules and test results of trained models on collected historical data：
·Building up the algorithm base and basic rules for algorithm primary screening. There are many algorithms for model training， e.g. Artificial Neural Network， Support Vector Machine， Gaussian Processes， etc. The definition of basic  rules should consider not only feature of algorithms but also the characteristics of data sources. For different available data sources， the primary selected algorithms can be different.
·Training models based on primary selected algorithms to be applicable for wind farm diagnosis and prognosis. The inputs of each model are historical data， and the output is the health condition of target wind turbine， which component is defective (if wind turbine is defective) and the fault information like location， type and severity.
·Verifying the effectiveness (e.g. detection rate) of the trained models using the collected historical data.
·Selecting the most proper algorithm for current data source according to validation results.
·Note that this step should be done off-line instead of on-line.
Step 203： Updatina models trained bv selected algorithm based on newly  collected online data consistently：
·Setting the models trained by selected algorithm as the initial diagnosis/prognosis models. Run these models for a certain period.
·Updating the initial wind turbine diagnosis/prognosis models based on newly collected online data. The inputs of the models are the newly available data and the output is the health condition of target wind turbine， which component is defective (if wind turbine is defective) and the fault information like location， type and severity.
·Verifying the effectiveness (e.g. detection rate) of the models to confirm the performance
·Note that this step should be done on-line.
Step 204： Using the updated models  to implement wind turhine  diagnosis/proanosis：
·Inputting real time newly coming monitoring data to the trained model.
·Running the model to implement the embedded algorithm.
·Outputting the health index of the detected wind turbine with information incl. whether the wind turbine is normal or defective， if defective which  component (s) is defective one (s) and the failure information like fault location， type and severity.
·Note that this step should be done on-line.
From the description above it can be obviously seen that the whole technology can properly select algorithm due to available data thus is flexible on data source， and can automatically adjust diagnosis/prognosis models by wind farm/turbine condition. Therefore， this self-adaptive CMS solution can provide wind farm a flexible， complete and accurate condition monitoring service without any hardware modification.
The term “on-line” refers to online real time monitor. The term “off-line” refers to offline models training.
Fig. 3 shows a structural module drawing of a wind farm condition monitoring system， comprising the following modules：
historical data acquiring module 301： used for acquiring historical SCADA data， and/or historical sensor data received from sensors which are installed in a wind farm， and historical wind turbine status which indicate wind turbines historical status in the wind farm， wherein the historical SCADA data covers operation data of the wind farm；
model training module 302： used for selecting algorithm for all types of historical data sources according to pre-defined basic rules， wherein the historical data sources include the historical SCADA data and/or the historical sensor data， training different models for different historical data sources using the algorithm selected for different types of historical data sources to establish relationship between the historical data sources and the historical wind turbine status；
model diagnosing module 303： used for acquiring real time SCADA data and/or real time sensor data as real time data source， selecting the trained model depending on type of the real time data source， inputting the real time data source to the selected trained model， obtaining the real time wind turbine status， which estimate wind turbines real time status in the wind farm， from the selected trained model.
The above-identified embodiments are only used for representing several  examples of the present invention， which are illustrated in detail， but shall not be understood to limit the protection scope of the present patent. It should be noted that， several modifications and/or improvements may be made for the skilled in the art， without going beyond the technical concept of the present invention， all of which fall into the protection scope of the present invention. Therefore， the protection scope of the present invention is dependent on the accompanied Claims.

Claims (18)

  1. A wind farm condition monitoring method， comprising the following steps：
    historical data acquiring step： acquiring historical SCADA data， and/or historical sensor data received from sensors which are installed in a wind farm， acquiring historical wind turbine status which indicate wind turbines historical status in the wind farm， wherein the historical SCADA data covers operation data of the wind farm；
    model training step： selecting algorithm for all types of historical data sources according to pre-defined basic rules， wherein the historical data sources include the historical SCADA data and/or the historical sensor data， training different models for different historical data sources using the algorithm selected for different types of historical data sources to establish relationship between the historical data sources and the historical wind turbine status；
    model diagnosing step： acquiring real time SCADA data and/or real time sensor data as real time data source， selecting the trained model depending on type of the real time data source， inputting the real time data source to the selected trained model， obtaining the real time wind turbine status from the selected trained model.
  2. The method according to claim 1， wherein the historical data acquiring step further comprises：
    verifying the wind turbine status to identifywhich data section of the historical SCADA data and/or the historical sensor data is in normal status and which data section of the historical SCADA data or the historical sensor data is in defective status.
  3. The method according to claim 1， wherein the model training step comprises：
    building up the algorithm base and basic rules for algorithm primary screening；
    selecting a primary algorithm from the algorithm base for each historical data source according to the basic rules；
    training models for each historical data source using the primary algorithm to establish relationship between the historical data sources and the historical wind  turbine status， wherein the inputs of each model are the historical data sources， and the outputs of each model are the historical wind turbine status；
    verifying the effectiveness of the trained models using the historical data source and the historical wind turbine report；
    selecting the most effective trained model from the models having same type of historical data source， selecting the primary algorithm corresponding to the most effective trained model as the algorithm selected for the type of historical data source， retraining models for each historical data source using the algorithm selected for the type of historical data source.
  4. The method according to claim 3， wherein the model diagnosing step comprises：
    acquiring real time SCADA data and/or real time sensor data as real time data source；
    selecting the trained model depending on type of the real time data source；
    inputting the real time data source to the trained model；
    running the trained model to implement the algorithm；
    obtaining the real time wind turbine report from the trained model.
  5. The method according to claim 1， wherein further comprises：
    model updating step： updating the trained models with the real time data source and the real time wind turbine status.
  6. The method according to claim 5， wherein the model updating step comprises：
    running the trained models for a predetermined period；
    retraining the trained models with the real time data source and the real time wind turbine report， wherein the inputs of each model are the real time data sources， the outputs of each model are the real time wind turbine status， and the algorithm used is algorithm selected for the type of real time data source.
  7. The method according to any one of claim 1 to 6， wherein the wind turbine status covers：
    health condition of wind turbine in the wind farm diagnosed as normal or defective status， and defective component with corresponding fault reason if health  condition of wind turbine is diagnosed as defective status.
  8. The method according to claim 7， wherein further comp rises：
    health alarming step： activating an alarm device when the health condition of the real time wind turbine status is diagnosed as defective status.
  9. A computer program comprising computer program code adapted to perform all of the steps of any one of the preceding claims when running on a computer.
  10. A computer program according to claim 9， which the computer program is embodied on a computer-readable medium.
  11. A wind farm condition monitoring system， comprising the following modules：
    historical data acquiring module： used for acquiring historical SCADA data， and/or historical sensor data received from sensors which are installed in a wind farm， acquiring historical wind turbine status which indicate wind turbines historical status in the wind farm， wherein the historical SCADA data covers operation data of the wind farm；
    model training module： used for selecting algorithm for all types of historical data sources according to pre-defined basic rules， wherein the historical data sources include the historical SCADA data and/or the historical sensor data， training different models for different historical data sources using the algorithm selected for different types of historical data sources to establish relationship between the historical data sources and the historical wind turbine status；
    model diagnosing module： used for acquiring real time SCADA data and/or real time sensor data as real time data source， selecting the trained model depending on type of the real time data source， inputting the real time data source to the selected trained model， obtaining the real time wind turbine status， which estimate wind turbines real time status in the wind farm， from the selected trained model.
  12. The system according to claim 11， wherein the historical data acquiring module further comprises：
    module used for verifying the wind turbine status to identify which data section of the historical SCADA data and/or the historical sensor data is in normal status and which data section of the historical SCADA data or the historical sensor  data is in detective status.
  13. The system according to claim 11， wherein the model training module comprises：
    module used for building up the algo rithm base and basic rules for algonthm primary screening；
    module used for selecting a primary algorithm from the algorithm basefor each historical data source according to the basic rules；
    mod ule used for training models for each historical data source using the primary algorithm to establish relationship between the historical data sources and the historical wind turbine status， wherein the inputs of each model are the historical data sources， and the outputs of each model are the historical wind turbine status；
    module used for verifying the effectiveness of the trained models using the historical data source and the historical wind turbine report；
    module used for selecting the most efffective trained model from the models having same type of historical data source， selecting the primary algonthm corresponding to the most effective trained model as the algorithm selected for the type of historical data source， retraining models for each historical data source using the algorithm selected for the type of historical data source.
  14. The system according to claim 13， wherain the model diagnosing module comprises：
    module used for acquiring real time SCADA data and/or real time sensor data as real time data source；
    module used for selecting the trained model depending on type of the real time data source；
    module used for inputting the real time data source to the trained model；
    module used for running the trained model to implement the algorithm；
    module used for obtaining the real time wind turbine report from the trained model.
  15. The system according to claim 11， wherein further comprises：
    model updating module： updating the trained models with the real time data source and the real time wind turbine status.
  16. The system according to ciaim 15， where the model updating module comprises：
    module used for running the trained models for a predetermined period；
    module used for retraining the trained models with the real time data source and the real time wind turbine report， wherein the inputs of each model are the real time data sources， the outputs of each model are the real time wind turbine status， and the algorithm used is algorithm selected for the type of real time data source.
  17. The system according to any one of claim 11 to 16， wherein the wind turbine status covers：
    health condition of wind turbine in the wind farm diagnosed as normal or defective status， and defective component with corresponding fault reason if health condition of wind turbine is diagnosed as defective status.
  18. The system according to claim 17， wherein further comprises：
    health alarming module： used for activating an alarm device when the health condition of the real time wind turbine status is diagnosed as defective status.
PCT/CN2014/092813 2014-12-02 2014-12-02 Wind farm condition monitoring method and system WO2016086360A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2014/092813 WO2016086360A1 (en) 2014-12-02 2014-12-02 Wind farm condition monitoring method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2014/092813 WO2016086360A1 (en) 2014-12-02 2014-12-02 Wind farm condition monitoring method and system

Publications (1)

Publication Number Publication Date
WO2016086360A1 true WO2016086360A1 (en) 2016-06-09

Family

ID=56090809

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/092813 WO2016086360A1 (en) 2014-12-02 2014-12-02 Wind farm condition monitoring method and system

Country Status (1)

Country Link
WO (1) WO2016086360A1 (en)

Cited By (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2555010A (en) * 2016-10-17 2018-04-18 Romax Tech Limited Determining loads on a wind turbine
WO2018076236A1 (en) * 2016-10-27 2018-05-03 Abb Schweiz Ag Wind farm and method and controller for feeding the same
EP3388910A1 (en) * 2017-04-10 2018-10-17 ABB Schweiz AG Method and apparatus for monitoring the condition of subsystems within a renewable generation plant or microgrid
CN109715936A (en) * 2016-09-13 2019-05-03 福斯4X股份有限公司 For monitoring the method and apparatus and computer program product of the state of at least one wind turbine
CN110094295A (en) * 2019-01-11 2019-08-06 北京京运通科技股份有限公司 Wind-powered electricity generation Centralized Monitoring management method and device
CN110488828A (en) * 2019-08-29 2019-11-22 南京智慧光信息科技研究院有限公司 Navigation lamp control method and robot system based on big data and artificial intelligence
CN110488803A (en) * 2019-08-29 2019-11-22 南京智慧光信息科技研究院有限公司 Braking method and robot system based on big data and artificial intelligence
CN110502015A (en) * 2019-08-29 2019-11-26 南京智慧光信息科技研究院有限公司 Method for control speed and robot system based on big data and artificial intelligence
CN113670790A (en) * 2021-07-30 2021-11-19 深圳市中金岭南有色金属股份有限公司凡口铅锌矿 Method and device for determining working state of ceramic filter
CN113884705A (en) * 2021-09-28 2022-01-04 上海电气风电集团股份有限公司 Monitoring method and system of cluster fan anemometer and computer readable storage medium
CN113982850A (en) * 2021-09-10 2022-01-28 大唐国信滨海海上风力发电有限公司 Fan comprehensive health analysis method and system fusing high-frequency and low-frequency signals
CN114320773A (en) * 2021-12-22 2022-04-12 中国大唐集团新能源科学技术研究院有限公司 Wind turbine generator fault early warning method based on power curve analysis and neural network
WO2022081230A1 (en) * 2020-10-14 2022-04-21 Microsoft Technology Licensing, Llc Autonomous control of supervisory setpoints using artificial intelligence
CN114562431A (en) * 2022-02-28 2022-05-31 北京金风慧能技术有限公司 Method and device for monitoring humidity of wind generating set
CN114856941A (en) * 2022-05-18 2022-08-05 江苏科技大学 Offshore wind power plant and unit fault diagnosis operation and maintenance system and diagnosis operation and maintenance method thereof
EP4151852A1 (en) * 2021-09-17 2023-03-22 Vestas Wind Systems A/S Determining an action to allow resumption wind turbine operation after a stoppage
CN116704735A (en) * 2023-08-08 2023-09-05 湖南江河能源科技股份有限公司 Hydropower station intelligent alarm method, system, terminal and storage medium
CN117972547A (en) * 2024-03-26 2024-05-03 华电电力科学研究院有限公司 Fault early warning method, device and medium for key components of wind turbine generator
WO2024212721A1 (en) * 2023-04-13 2024-10-17 河北建投新能源有限公司 Wind power big data analysis method and system based on cloud computing

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101872165A (en) * 2010-06-13 2010-10-27 西安交通大学 Method for fault diagnosis of wind turbines on basis of genetic neural network
US20110004356A1 (en) * 2007-12-19 2011-01-06 Vestas Wind Systems A/S Event-based control system for wind turbine generators
CN102944416A (en) * 2012-12-06 2013-02-27 南京匹瑞电气科技有限公司 Multi-sensor signal fusion technology-based fault diagnosis method for wind turbine blades
CN103323772A (en) * 2012-03-21 2013-09-25 北京光耀能源技术股份有限公司 Wind driven generator operation state analyzing method based on neural network model
CN103758696A (en) * 2014-01-09 2014-04-30 华北电力大学 SCADA (supervisory control and data acquisition) temperature parameter based wind turbine set security evaluation method
CN103835882A (en) * 2014-03-18 2014-06-04 上海电机学院 Large wind generating set state monitoring and fault diagnosis system

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110004356A1 (en) * 2007-12-19 2011-01-06 Vestas Wind Systems A/S Event-based control system for wind turbine generators
CN101872165A (en) * 2010-06-13 2010-10-27 西安交通大学 Method for fault diagnosis of wind turbines on basis of genetic neural network
CN103323772A (en) * 2012-03-21 2013-09-25 北京光耀能源技术股份有限公司 Wind driven generator operation state analyzing method based on neural network model
CN102944416A (en) * 2012-12-06 2013-02-27 南京匹瑞电气科技有限公司 Multi-sensor signal fusion technology-based fault diagnosis method for wind turbine blades
CN103758696A (en) * 2014-01-09 2014-04-30 华北电力大学 SCADA (supervisory control and data acquisition) temperature parameter based wind turbine set security evaluation method
CN103835882A (en) * 2014-03-18 2014-06-04 上海电机学院 Large wind generating set state monitoring and fault diagnosis system

Cited By (31)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109715936A (en) * 2016-09-13 2019-05-03 福斯4X股份有限公司 For monitoring the method and apparatus and computer program product of the state of at least one wind turbine
GB2555010B (en) * 2016-10-17 2019-09-25 Romax Tech Limited Determining loads on a wind turbine
GB2555010A (en) * 2016-10-17 2018-04-18 Romax Tech Limited Determining loads on a wind turbine
WO2018076236A1 (en) * 2016-10-27 2018-05-03 Abb Schweiz Ag Wind farm and method and controller for feeding the same
EP3388910A1 (en) * 2017-04-10 2018-10-17 ABB Schweiz AG Method and apparatus for monitoring the condition of subsystems within a renewable generation plant or microgrid
WO2018188774A1 (en) * 2017-04-10 2018-10-18 Abb Schweiz Ag Method and apparatus for monitoring the condition of subsystems within a renewable generation plant or microgrid
CN110709789B (en) * 2017-04-10 2024-05-31 Abb瑞士股份有限公司 Method and apparatus for monitoring the condition of subsystems within a renewable power generation device or microgrid
US11604461B2 (en) 2017-04-10 2023-03-14 Hitachi Energy Switzerland Ag Method and apparatus for monitoring the condition of subsystems within a renewable generation plant or microgrid
CN110709789A (en) * 2017-04-10 2020-01-17 Abb瑞士股份有限公司 Method and apparatus for monitoring the condition of a subsystem within a renewable power generation device or microgrid
CN110094295B (en) * 2019-01-11 2020-05-08 北京京运通科技股份有限公司 Wind power centralized monitoring management method and device
CN110094295A (en) * 2019-01-11 2019-08-06 北京京运通科技股份有限公司 Wind-powered electricity generation Centralized Monitoring management method and device
CN110488828A (en) * 2019-08-29 2019-11-22 南京智慧光信息科技研究院有限公司 Navigation lamp control method and robot system based on big data and artificial intelligence
CN110488828B (en) * 2019-08-29 2023-01-03 南京智慧光信息科技研究院有限公司 Navigation light control method based on big data and artificial intelligence and robot system
CN110488803A (en) * 2019-08-29 2019-11-22 南京智慧光信息科技研究院有限公司 Braking method and robot system based on big data and artificial intelligence
CN110502015A (en) * 2019-08-29 2019-11-26 南京智慧光信息科技研究院有限公司 Method for control speed and robot system based on big data and artificial intelligence
CN110488803B (en) * 2019-08-29 2023-01-03 南京智慧光信息科技研究院有限公司 Braking method based on big data and artificial intelligence and robot system
WO2022081230A1 (en) * 2020-10-14 2022-04-21 Microsoft Technology Licensing, Llc Autonomous control of supervisory setpoints using artificial intelligence
CN113670790A (en) * 2021-07-30 2021-11-19 深圳市中金岭南有色金属股份有限公司凡口铅锌矿 Method and device for determining working state of ceramic filter
CN113670790B (en) * 2021-07-30 2024-03-22 深圳市中金岭南有色金属股份有限公司凡口铅锌矿 Method and device for determining working state of ceramic filter
CN113982850B (en) * 2021-09-10 2023-11-10 大唐国信滨海海上风力发电有限公司 Fan comprehensive health analysis method and system integrating high-low frequency signals
CN113982850A (en) * 2021-09-10 2022-01-28 大唐国信滨海海上风力发电有限公司 Fan comprehensive health analysis method and system fusing high-frequency and low-frequency signals
EP4151852A1 (en) * 2021-09-17 2023-03-22 Vestas Wind Systems A/S Determining an action to allow resumption wind turbine operation after a stoppage
CN113884705A (en) * 2021-09-28 2022-01-04 上海电气风电集团股份有限公司 Monitoring method and system of cluster fan anemometer and computer readable storage medium
CN114320773A (en) * 2021-12-22 2022-04-12 中国大唐集团新能源科学技术研究院有限公司 Wind turbine generator fault early warning method based on power curve analysis and neural network
CN114320773B (en) * 2021-12-22 2023-09-22 大唐可再生能源试验研究院有限公司 Wind turbine generator system fault early warning method based on power curve analysis and neural network
CN114562431A (en) * 2022-02-28 2022-05-31 北京金风慧能技术有限公司 Method and device for monitoring humidity of wind generating set
CN114856941A (en) * 2022-05-18 2022-08-05 江苏科技大学 Offshore wind power plant and unit fault diagnosis operation and maintenance system and diagnosis operation and maintenance method thereof
WO2024212721A1 (en) * 2023-04-13 2024-10-17 河北建投新能源有限公司 Wind power big data analysis method and system based on cloud computing
CN116704735B (en) * 2023-08-08 2023-11-03 湖南江河能源科技股份有限公司 Hydropower station intelligent alarm method, system, terminal and storage medium
CN116704735A (en) * 2023-08-08 2023-09-05 湖南江河能源科技股份有限公司 Hydropower station intelligent alarm method, system, terminal and storage medium
CN117972547A (en) * 2024-03-26 2024-05-03 华电电力科学研究院有限公司 Fault early warning method, device and medium for key components of wind turbine generator

Similar Documents

Publication Publication Date Title
WO2016086360A1 (en) Wind farm condition monitoring method and system
EP3221579B1 (en) Wind turbine condition monitoring method and system
EP3638900B1 (en) Independent monitoring system for a wind turbine
Zhang et al. Wind turbine fault detection based on SCADA data analysis using ANN
US9141915B2 (en) Method and apparatus for deriving diagnostic data about a technical system
JP2021064370A (en) Method and system for semi-supervised deep abnormality detection for large-scale industrial monitoring system based on time-series data utilizing digital twin simulation data
US11761427B2 (en) Method and system for building prescriptive analytics to prevent wind turbine failures
KR20170031985A (en) Fault detection and diagnostics method of air-conditioning system
CN107729985B (en) Method for detecting process anomalies in a technical installation and corresponding diagnostic system
CN110709789B (en) Method and apparatus for monitoring the condition of subsystems within a renewable power generation device or microgrid
US20210397950A1 (en) Abnormal driving state determination device and method using neural network model
CA3035871A1 (en) Method and device for monitoring a status of at least one wind turbine and computer program product
Sharma et al. Condition monitoring of wind turbines: a review
Tchakoua et al. New trends and future challenges for wind turbines condition monitoring
Mazidi et al. Performance analysis and anomaly detection in wind turbines based on neural networks and principal component analysis
Zhang Comparison of data-driven and model-based methodologies of wind turbine fault detection with SCADA data
US11339763B2 (en) Method for windmill farm monitoring
CN118395126A (en) Fault diagnosis method, system and storage medium for equipment rotation state analysis
José et al. Improvements in failure detection of DAMADICS control valve using neural networks
Dinh et al. Implementation of Digital Twin-Assisted Condition Monitoring and Fault Diagnosis for Wind Turbines
Blesa et al. Fault diagnosis of a wind farm using interval parity equations
CN114638060A (en) Fault prediction method, system, electronic device and computer program product
CN118137926B (en) Frequency modulation control method for three-phase asynchronous motor based on electromechanical signal analysis
CN115342036B (en) Abnormal early warning method and system for variable pitch motor of wind power generation assembly machine
CN118101429B (en) Power distribution network communication fault diagnosis method and system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 14907331

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 14907331

Country of ref document: EP

Kind code of ref document: A1