| Title |
Time-Delay Neural Network based P&O MPPT Control Method |
| Authors |
Hak-Soo Kim ; Yong-Kyo Seo ; Sung-Kwan Kang ; Dong-Hyun Lim ; Eui-Cheol Nho |
| DOI |
https://doi.org/10.6113/TKPE.2025.30.5.450 |
| Keywords |
Maximum power point tracking(MPPT); PV Operating Point(PV-OP) Estimator; Feedforward Neural Network(FNN); Time Delay Neural Network(TDNN) |
| Abstract |
This study proposes a new time-delay neural network (TDNN)-based perturb and observe maximum power point tracking (MPPT) control method. The conventional NN-based PV operating point (PV-OP) estimator provides precise PV-OP estimation to improve MPPT performance, but it requires five sensors for irradiance, PV output voltage, PV output current, converter output voltage, and converter output current. A new concept is proposed in this study to reduce the number of sensors from 5 to 2. The reduced number of sensors decreases the input feature number of the PV-OP estimator. Although the number of input features is reduced, the number of learning datasets is increased by the incorporation of various duty step sizes. The enlarged dataset and TDNN structure can guarantee high MPPT performance. The usefulness of the proposed method is verified through simulations and experiments. |