Hybrid Flower Pollination and Recurrent Neural Network (Fprnn) For Forecasting OPEC Annual Petroleum Consumption

The development and use of intelligent methods as an emerging trend is attracting attention in the academia and industries, including petroleum industry. Forecasting OPEC petroleum consumption can go to a greater extend in helping the management of OPEC to properly manage petroleum demand and supply for its member countries, investment analysis, research related to energy as well as the analysis of revenue generation. Previous research designed and implemented flower pollination optimized neural network (FPNN) for forecasting OPEC annual petroleum consumption. However, NN is essentially designed for multidimensional data whose attributes are widely independent on one another. As an improvement over FPNN, this paper proposed and implemented a hybrid optimized technique for forecasting OPEC annual petroleum consumption using flower pollination and recurrent neural network (FPRNN). An open-source dataset was obtained from the official website of U.S energy information administration. The proposed FPRNN is applied to predict OPEC annual consumption of petroleum from 2009 to 2018. The performance of the proposed FPRNN is evaluated using FPNN. The results showed that the proposed FPRNN outperformed FPNN. Thus, our proposed method has better performance to OPEC annual petroleum consumption prediction