| Wind power forecasting and electricity storage are becoming increasingly important in energy systems with high shares of fluctuating wind power.The Power to Gas(P2G)technology is one promising option for solving the challenge of long-term electricity storage.With the development of this technology,electric power and natural gas system can be coupled into integrated electricity gas system(IEGS)to improve wind power utilization efficiency.This paper firstly introduces the factors influencing wind power,the methods for wind power forecasting and the indexes to evaluate wind power forecasting.By introducing firefly behavior strategy and Lévy flight strategy,the basic teaching learning based optimization(TLBO)can be transformed to the improved teaching learning based optimization(ITLBO)whose superiority is proved by implementing it on a range of standard benchmark functions.The support vector regression(SVR)whose parameters are optimized by ITLBO algorithm is used to forecast short-term wind power.The Grid method,TLBO algorithm and ITLBO algorithm are used for parameters optimization respectively and compared by the prediction results.The simulation results show that ITLBO algorithm can enhance the reliability and precision of the prediction system.Based on the model of IEGS,the improved 9-node power system and 7-node natural gas system as a case of economic scheduling solved by TLBO algorithm is used to analyze the role of P2 G units to absorb wind power in IEGS.Case studies demonstrate that the wind power can reduce system cost and carbon dioxide emission.P2 G units can reduce the level of wind abandonment and enhance the effect of wind power to reduce carbon emission.And the improving the capacity of P2 G units is effective mean to improve the role to absorb wind power.At the same time,the synthetic natural gas produced by P2 G units can reduce the gas-fired power plants cost,increase its output and coordinate the operation of the gas turbine unit. |
