Title :
A neural network controller for maximum power point tracking with 8-bit microcontroller
Author_Institution :
Dept. of Mech. Eng., Ming Chi Univ. of Technol., Taipei, Taiwan
Abstract :
This paper presents the implementation of a neural network controller (NNC) for maximum power point tracking (MPPT). A fuzzy logic controller (FLC) with three fuzzy sets is designed and its look-up control table is used to train a 3-layer feed-forward neural network. The MPPT controllers are implemented with an 8-bit microcontroller, PIC16F877A, and tested on a power source which comprises a voltage source and a series resistor. Experimental results show that the implemented NNC can stably extract the maximum power. The performance of the NNC approximates to the performance of the real-time FLC and the lookup table FLC. The RAM used in the NNC is less than the RAM used in the real-time FLC. And, the ROM used in the NNC is less than the ROM used in the lookup table FLC.
Keywords :
fuzzy control; fuzzy set theory; maximum power point trackers; microcontrollers; neurocontrollers; random-access storage; read-only storage; 8-bit microcontroller; FLC; MPPT controllers; NNC; PIC16F877A; RAM; ROM; fuzzy logic controller; fuzzy sets; look-up control table; maximum power point tracking; neural network controller; Artificial neural networks; Microcontrollers; Neurons; Pulse width modulation; Random access memory; Real time systems; Zirconium; fuzzy logic controller; maximum power tracking; neural network;
Conference_Titel :
Industrial Electronics and Applications (ICIEA), 2011 6th IEEE Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4244-8754-7
Electronic_ISBN :
pending
DOI :
10.1109/ICIEA.2011.5975718