• State Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 1 answer
  • Subscribers 85 subscribers
  • Views 896 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

EDGE-AI-STUDIO: Edge AI studio Model Composer - Arc fault data

Cemil Karaköz
Prodigy 50 points
Part Number: EDGE-AI-STUDIO
Other Parts Discussed in Thread: TMDSCNCD28P55X, TIEVM-ARC-AFE, TIDA-010955, SYSCONFIG

Tool/software:

I would like to create more data with existing PV systems, but I have a couple questions about the arc fault data:

  1. What does the 300nF refer to in the CSV file names?
  2. What is the difference between files that are used for training, testing and validating?
  3. What difference does the data between a 4th order filter and 8th order filter do for the data, and which one should I choose?
  4. What does range mean in the graphs and can I change the range in case I want to use an externall current transformers?
  5. What is the large peak in the frequency domain, both in the normal and arc files?

Thank you and with kind regards,

Cemil

  • 0
    TI__Guru**** 474421 points

    Hello,

    What device are you using?

    Thanks

    ki

  • 0 in reply to Ki
    Prodigy 50 points

    Hi,

    I am currently using the TIEVM-ARC-AFE in combination with the TMDSCNCD28P55X.

    With kind regards,

    Cemil

  • 0 in reply to Cemil Karaköz
    TI__Guru**** 474421 points

    Thank you. I will let the device experts know of this thread.

  • 0
    TI__Expert 5005 points

    Hi Cemil

    Apology for the delay in response.
    Can you send screen shot of below questions for better clarity?
    1. the files name 
    2. the range in the graph

    3.large peak in the frequency domain.

    Thank you

    Amir Hussain

  • 0 in reply to Amir Hussain
    Prodigy 50 points

    Hi Amir,

    More background info: i am using Edge AI studio in my browser. In here I created a example project using the "arc_fault_example_dsk" data set. These files are (probably) the same as the ones found in the TIDA-010955 project, but the dataset has less files then the one found in the TIDA-010955. I also haven't tried to edit the files yet so I am unsure if the content within these .csv files are different, but they serve the same function. Edge AI studio can be accessed and the data can be viewed without a devboard.

    Answer:

    I answered your questions in the picture.

    Thank you.

    Cemil

  • 0
    TI__Prodigy 370 points

    Hi Cemil,

    Regarding "What is the difference between files that are used for training, testing and validating?"

    Training files are like notebooks/knowledge that the model uses to learn

    Validation files are like internal assessments --> used to test the model's learning at the end of each epoch.

    Test Files are like semester end exams --> they are used to report the final score & metrics of the model

  • +1
    TI__Expert 5005 points

    Hello Ki

    1. 300nf is capacitor used in decoupling network of the arc fault test bed. You can find more about it the UL standard.

    3. This depends on the system requirements. A 8th order filter attenuates the out off band noise more, while a 4th order filter can be implemented at lower cost. We have both options on the board for evaluation. If performance with 4th order is sufficient i would recommend this one.

    4. Ideal the range should not appear in the frequency domain. We working to improve it

    5. This peak is corresponding to switching harmonic of the inverter.

    Thank you

    Amir Hussain

  • 0 in reply to Amir Hussain
    Prodigy 50 points

    Hello Amir,

    I have one more question, is it possible to get the sysconfig file that is used in this project (TIDA-010955).

  • 0 in reply to Cemil Karaköz
    TI__Expert 5005 points

    Hi Cemil

    Unfortunately, the sysconfig file is not available in the released version of firmware. 

    Thank you

    Amir Hussain