Supervised Machine Learning in Wind Forecasting and Ramp Event Prediction

Cover
SUPERVISED MACHINE LEARNING IN WIND FORECASTING AND RAMP EVENT PREDICTION
Copyright
Dedication
Contents
List of figures
List of tables
Biography
	About the series editor
	About the authors
Preface
Acknowledgments
Acronyms
1 Introduction
	1.1 Renewable energy focus
	1.2 Wind energy: issues and challenges
	1.3 Machine learning in allied areas of wind energy
	1.4 Scope and outline of the book
	References
2 Wind energy fundamentals
	2.1 Basics of wind power
	2.2 Wind resource assessment
	2.3 Wind speed distribution
		2.3.1 Probability density functions for wind speed
			2.3.1.1 Weibull distribution
			Parameter estimation using maximum log likelihood
			2.3.1.2 Lindley distribution for wind speed
			Parameter estimation for Lindley PDF
			2.3.1.3 Stacy-Mihram distribution model
	2.4 Wind turbine micrositing
	References
3 Paradigms in wind forecasting
	3.1 Introduction to time series
	3.2 Wind forecasting: overview
		3.2.1 Techniques based on prediction horizon
		3.2.2 Techniques based on methods of forecasting
	3.3 Statistical methods
		3.3.1 ARMA model
		3.3.2 ARIMA model
	3.4 Machine learning-based models
	3.5 Hybrid wind forecasting methods
	References
4 Supervised machine learning models based on support vector regression
	4.1 Support vector regression
	4.2 ε-support vector regression
		Stock price prediction: a case study
	4.3 Least-square support vector regression
		Power generation prediction: a case study
	4.4 Twin support vector regression
		Wine quality prediction: a case study
	4.5 ε-twin support vector regression
		Kernel ε-twin support vector regression
		River flow forecasting: a case study for Nile river
	References
5 Decision tree ensemble-based regression models
	5.1 Random forest regression
		5.1.1 Monthly rainfall prediction: a case study
		5.1.2 Crude oil price prediction: a case study
	5.2 Gradient boosted machines
		5.2.1 Monthly rainfall prediction: a case study
		5.2.2 Crude oil price prediction: a case study
	References
6 Hybrid machine intelligent wind speed forecasting models
	6.1 Introduction
	6.2 Wavelet transform
	6.3 Framework of hybrid forecasting
		Description of datasets
	6.4 Results and discussion
	6.5 Empirical mode decomposition-based SVR variants for wind speed prediction
	References
7 Ramp prediction in wind farms
	7.1 Ramp events in scientific and engineering activities
		7.1.1 Reservoir wall: a case study from Valsad, Gujarat
		7.1.2 Forest risk management: a case study
		7.1.3 Strawberry cultivation case study, Monterey, CA
		7.1.4 Air quality: a case study for Bogota, Colombia
	7.2 Ramp events in wind farms
		7.2.1 Ramp event error analysis for ε-SVR and LS-SVR
		7.2.2 Ramp event error analysis for TSVR and ε-TSVR
		7.2.3 Case study for ramp event analysis
	7.3 Ramp event analysis for onshore and offshore wind farms
		7.3.1 Case study for onshore and offshore wind farms
		7.3.2 Discussion on uncertainties in ramp events
	References
8 Supervised learning for forecasting in presence of wind wakes
	8.1 Introduction
	8.2 Wind wakes
		8.2.1 Jensen's and Frandsen's single wake models
		8.2.2 Proposed model for wind wakes
		8.2.3 Case study for single-wake model
		8.2.4 Multiple-wake model
	8.3 Wake effect in wind forecasting
		Grey correlation analysis
	8.4 Results
		8.4.1 Forecasting results for five-turbine wind farm layout
		8.4.2 Forecasting results for 15-turbine wind farm
	References
Epilogue
	References
APPENDIX
A Introduction to R for machine learning regression
	A.1 Data handling in R
		A.1.1 Data handling via vectors, lists, and data frames
		A.1.2 Importing data through .xlsx and .csv files
		A.1.3 Line plots, histograms, and autocorrelation function for wind speed in R
		A.1.4 Parameter estimation for wind speed probability density functions in R
	A.2 Linear regression analysis in R
	A.3 Support vector regression in R
	A.4 Random forest regression in R
	A.5 Gradient boosted machines in R
Index