A Framework for Forecasting Wind Speed and Power Using Adaboost with Back Propagation

Electricity can be generated by a variety of ways. Wind power has many characteristics which other fossil energy does not have, such as clean, intermittent and randomness. This is because the wind is a natural phenomenon. Wind energy converts into mechanical energy in the way that wind blow through fans to drive rotor rotation. The reason why the demand for wind power around the world grows involves many aspects, including the shortage of energy, change in climate, the progress of economy and technology, etc. Due to wind is intermittent and less dispatchable, wind power fluctuates as the wind fluctuating and is uncontrollable. The way to solve the problem is forecasting the wind power. The wind speed and wind power are considered as the main input to forecasting models. The new forecasting model is adaboost with Back propagation NN. The new model mainly focusing the speed and accuracy. The new algorithm adaboost with Back propagation NN will improve the accuracy of the forecasting power with the forecasting wind. This is because using the forecasting wind speed instead of the original one can reduce the error which is caused by the training data and the noise which is produced in sampling process or data transmission.