The need of accurate wind turbine power curve modelling is essential as it provides great insight on the performance of the power system and ideal estimate of generation of power by turbines. However, presence of non linear relationship between output power of turbine and its primary and secondary parameters imposes restrictions to predict exact power generated. In this work wind power curve modelling is accomplished using different robust linear techniques that reduces the effects of outliers and provides precise results in terms of power generation. Mean Absolute Error(MAE), Root Mean Square Error(RMSE), and R^2 score were used as a measure of approximation method accuracy. Since above metrics doesn't provide any information about fitted model insights like overfiiting and underfitting, so bias and variance are considered as the two important aspects that clearly assesses model complexity. By decomposing the MSE into bias and variance a clear insight about model structure is obtained with a potential way with regard to which error component is likely to contribute more degradation of model performance.