Enhancing cooling tower performance with condition monitoring and machine learning based drift detection

Sorry you have no rights to view this entry!

6 Comments

  1. tiziana.segreto

    Dear Sina Nahvi,
    thank you for your interesting presentation.
    I have a question regarding the input layer of the utilized ANN. In your presentation, 7 diverse sensorial data were detected.
    Did you use all these 7 data as an input to the ANN?

  2. Dear Tiziana Segreto,
    thanks for your question.
    As you mentioned, we collected data from 7 different sensors. For each dependent variable (i.e. all mentioned variables except wet bulb temperature and humidity, which are environmental data), a network with 6 input and 1 output was trained. In such a way, that each time, one of the dependent variables was used as output and the other 6 remaining variables were used as input.

  3. tiziana.segreto

    Thanks for your reply! I have another question regarding the hidden layer nodes. How were the number of hidden layer nodes chosen?

  4. Thanks for your reply!

Leave a Reply

Your email address will not be published. Required fields are marked *