|
| Identifying equivalent circuit of electrochemical impedance spectroscopy for zinc electrowinning based on deep learning |
|
Received:April 01, 2022
Revised:April 01, 2022
Accepted:April 08, 2022
Published Online:July 04, 2022
|
| View Full Text View/Add Comment Download reader |
| DOI： |
| KeyWord:EIS; equivalent circuit; identification; CNN-LSTM |
| Author | Institution |
| GAO Bo-bin |
Kunming University of Science and Technology,Kunming |
| YANG Chang-jiang |
Kunming University of Science and Technology,Kunming |
|
| Hits: 2048 |
| Download times: 941 |
| Abstract: |
| Electrochemical impedance spectroscopy (EIS) is widely used to study the electrochemical process of zinc electrowinning. It is generally analyzed by the equivalent circuit method, and the selection of correct equivalent circuit becomes the primary important. In this paper, a deep learning model is proposed to identify the equivalent circuit of EIS, which combines convolutional neural network ( CNN ) and long short-term memory neural network ( LSTM ). By collecting the equivalent circuit of typical zinc electrowinning process and the corresponding EIS data, the data are preprocessed and then entered into the convolutional neural network to extract the characteristic information of EIS. The stacked long and short-term memory network is used to extract the deep-level sequence characteristics associated with the sequence EIS data, and the selected equivalent circuit is obtained. The accuracy of selection of CNN-LSTM model reaches 90.87 %, which is at least 1.66 % higher than that of CNN, LSTM and corresponding improved models. The method was applied to the analysis of real zinc electrowinning EIS, and the model predicted probability distribution was consistent with the experimental test, indicating that the CNN-LSTM model can realize the efficient identification of equivalent circuit for the EIS of zinc electrowinning, which has important theoretical significance and technical value. |
| Close |
|
|
|