Improvements in Fitting Accuracy of Weibull Distribution for Wind Data by Capturing Monthly and Diurnal Variability in Wind Speeds

Existing approaches for determining wind speed distribution usually fit a single Weibull distribution to the entire measured data or fit multiple Weibull distributions to the data segregated by months. However, wind speed varies not only monthly or seasonally but also hourly. The present study captures monthly and diurnal variability by fitting a Weibull distribution to data measured at four locations in the central-western part of India. Two methods, a maximum likelihood (MLE) method and an empirical method, were used to estimate the Weibull parameters with several goodness-of-fit tests including R², root mean square error (RMSE), average power density error, and average energy output error. The findings demonstrated a marginal improvement in fitting accuracy with the implementation of monthly–hourly bin segmentation and fitting calculating Weibull parameters for each dataset.