Abstract:
This thesis proposes a method for evaluating the reliability of the electrical power system of Thailand by artificial neural networks (ANN), which is a multi-layer feed forward neural network. This method compares with adaptive neuro-fuzzy inference systems (ANFIS), which is a hybrid method of artificial neural network and fuzzy rules. Determine the priority list of generator, installed capacity and force outage rate are inputs. And expected energy not supplied index is output. The information entered into the learning comes from statistical and probabilistic calculations. It was found that the method of evaluating the reliability of electricity generation using artificial neural networks was better than adaptive neuro-fuzzy inference systems.
In addition, the proposed method of wind energy data simulation using the Weibull distribution, artificial neural networks and adaptive neuro-fuzzy inference systems. Determine the wind energy 2 month ago, wind energy 1 month ago and average wind speed are inputs. And wind energy is output. The information entered into the learning comes from the recorded value of wind power from the Promthep alternative energy station, Phuket, Thailand. It was found that artificial neural networks delivers the best results.