Neural network (NNs) approach for controlling switching power converters has been investigated in this paper. Several NNs techniques including the Back propagation training algorithm, Back propagation with momentum, Back propagation with adaptive learning rate and the Back propagation with Levenberg Marquardt optimisation have been investigated and the most later method has been selected for identification and control of Quasi-Resonant Converters (QRCs). A neural network emulator is designed in this paper, which reproduce the converter dynamic behaviour with great accuracy. The NN controller is developed and applied to regulate the converter output voltage. Results obtained from the NNC proved the validity and fast dynamic response of the proposed controller.
|Title of host publication||Proceedings of the 27th Annual Conference of the IEEE Industrial Electronics Society|
|Number of pages||6|
|Publication status||Published - 2001|
|Event||Proceedings of the 27th Annual Conference of the IEEE Industrial Electronics Society - Denver, CO, 29 Nov - 2 Dec, 2001|
Duration: 1 Jan 2001 → …
|Conference||Proceedings of the 27th Annual Conference of the IEEE Industrial Electronics Society|
|Period||1/01/01 → …|