---

  Storing the electrical energy in large scale is impossible. So, it is necessary to identify the factors affecting the electricity demand. Researchers have used different methods to forecast the future demand of electricity, among them intelligent methods and fuzzy based methods are more popular. Since ANFIS structure is based on researcher’s experience about phenomenon, the created structure may not have the best result. Therefore, we used PSO-ANFIS structure.

  In this paper long term electricity demand is forecasted until the year 2025 by hybrid PSO-ANFIS algorithm. The results confirm the high power of the Adaptive Neural based Fuzzy Inference System in forecasting the electricity demand. Results also indicate that the forecasted electricity demand will be 401 billion KWh in 2025. The prediction performance of the proposed technique is more accurate than the ARIMA model.

---

Price shocks lead to oil price volatility in world oil markets. In response to this volatility, economic growth may take different regime and behavior patterns in different situation. Investigating this multi behavior patterns can be useful for policymakers to reduce the effect of oil price volatility. In this study, an EGARCH model has developed using the seasonal data of OPEC oil basket nominal prices during 1367:Q1-1389:Q4. Markov switching models is also applied to investigate the multi behavior patterns of economic growth in response to oil price volatility in Iran. The results show that positive oil price shocks sharply lead to formation of oil price volatility, but, the negative price shocks will slightly reduce oil price volatility. Iranian economic growth is affected by this volatility under three different behavior regimes. If the economy switch to one of the regimes (low, medium, high economic growth), the probability of transition between these regimes and their duration is different. So, oil price volatility as a reason for low economic growth in Iran may cause the economy switch to its lower situation.