Wind-Speed-Prediction-using-Time-Series-Analysis

Wind Speed Prediction using LSTM

INTRODUCTION

Why Do We Need Wind Speed Prediction?

To predict the power/energy needed to meet the demand and supply equilibrium. Wind power is used as it is clean and Renewable energy Wind energy is uncertain and variable , so accurate forecasting model is important. Accurate prediction is important for allocation and planning of wind energy power plants.

The Data collected for the Wind Speed Prediction has been obtained from the ‘Meteoblue’ Weather Engine that records various weather parameters such as Temperature, Relative Humidity, Pressure, Precipitation, Solar Radiation, Wind Speed and Direction etc.

The Data collected was of Canton, Basel-City, Switzerland for the duration of 10 years (1 Jan 2001 to 31 Dec 2010) as well as data collected from NIWE (National Institute of Wind Energy)

LSTM Model for TIME SERIES PREDICTION

A single hidden layer ( 4 neuron ) LSTM model. Input data was accordingly modified to convert it into a step data that can be given as input to LSTM Network.