Adaptive neuro-fuzzy inference system based optimal spinning reserve identification in competitive electricity market

This paper proposes an adaptive neuro-fuzzy inference system based optimal spinning reserve identification (ANFIS-OSR) in competitive electricity market. The probability of generation outage or forced outage rate (FOR), the value of lost load (VOLL), the daily load forecast uncertainty (LFU) and the spinning reserve price are used as the input values. The spinning reserve requirement in percentage of total real power generation is the output of the proposed ANFIS-OSR. The method can evaluate several uncertain factors affecting spinning reserve requirements in a soft computing manner. The ANFIS-OSR is tested on the IEEE one area reliability test system (RTS-96) and compared to the largest unit real power generation, ten percent of total real power generation and the minimum expected reliability cost (ERC) condition. It is shown that the solution of ANFIS-OSR is very close to the minimum ERC condition with a much faster CPU time.