Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

The scope of the Wintell platform is to supervise optimally operations and health of wind mills. As a result, wind turbines availability and performance will be significantly improved. Advanced machine learning algorithms, called Smart agents, are monitoring several hundred parameters per turbine to continuously diagnose health and operational performance of the system. As soon as abnormal situations are detected, these are reported and end-users are notified. The system provides reporting facilities and alleviates the workflow of end-users. 

...

  1. Click Health Status tab  in Asset.
  2. Click New. 
  3. In General tab, put the Period (sec), defines the repeat interval in which the smart agent will run.
  4. Put Title, name of the agent.
  5. Detail the smart agent principle in Description, it contains the detailed information about the agent.
  6. Define the History source, eDNA per default. 
  7. Choose Status, the type of agent. The default type if Tag. The options are: Tag, Script, Datamaestro, Stored Script and External. 


    Agent TypeDescription
    TagIt consists in establishing a threshold value to a respective status (OPERATIONAL, WARNING, ALARM, SERVICE and FAILED). It is possible to connect one or more equipment with one or more tags, selected from a list. The equipment identification number is the one used for the CMMS - Senergy.
    ScriptIt consists in writing a JAVAscript using a script interface. The script can be tested before being implemented. 
    DatamaestroIt is based on a DATAmaestro Analytics predictive model.  
    Stored ScriptIt runs on wintell hub. Can perform simple computations or do the association with external libraries.
    ExternalIt runs outside the wintell hub. Push the results periodically to the WINTELL hub through a web API.



  8. The Definition tab changes depending on the agent type. 
  9. Define the History source, eDNA per default.

...

Create DATAmaestro agent

To create new DATAmaestro agent:  

Beforehand, the predictive model must be created in DATAmaestro Analytics.  
  1. Click Health status tab in Asset.
  2. Click New button.
  3. In General tab, put Period (in seconds).
  4. Define the agent scope in Description field - it sets the Agent’s name.
  5. Put History source, if other than eDNA.
  6. In Definition tab, put Equipment number. This field is filled automatically. 
  7. Put DATAmaestro Project name. Example: /wintell/XXX.dm.project
  8. Put Task name. It is the name of the model 
  9. Put status Thresholds.
    • Select Status: OPERATIONAL, WARNING, ALARM, SERVICE and FAIL. 

      StatusDescription
      OperationalEquipment is working appropriately
      WarningThere is a possible danger or problem in the near future
      AlarmIndication of fault
      ServiceEquipment is in maintenance
      FailedEquipment failed


    • Put a Threshold value.
    • Click Add button.
    • Click Remove Selected, to discard a Threshold.

Create Text agent

To create new Text agent:  


  1. Click Health status tab in Asset.
  2. Click + Create button.
  3. Select Text.
  4. Put Period (in seconds).
  5. Define the agent scope in Description field - agent’s name.
  6. Put History source, if other than eDNA.
  7. Write the text Script. Select tags in Tags. There is Help tab for JAVAscript tutorial with some important elements of the code.
  8. Click Save.

...