Data science for wind energy / / Yu Ding.

• VerfasserIn: Ding, Yu (Electrical and Computer Engineer),
• Place / Publishing House: Boca Raton : : CRC Press,, [2020]
• Year of Publication: 2020
• Edition: 1st ed.
• Language: English
• Physical Description: 1 online resource (425 pages)

Data Science for Wind Energy provides an in-depth discussion on how data science methods can improve decision making for wind energy applications, near-ground wind field analysis and forecast, turbine power curve fitting and performance analysis, turbine reliability assessment, and maintenance optimization for wind turbines and wind farms. A broad set of data science methods covered, including time series models, spatio-temporal analysis, kernel regression, decision trees, kNN, splines, Bayesian inference, and importance sampling. More importantly, the data science methods are described in the context of wind energy applications, with specific wind energy examples and case studies. Features Provides an integral treatment of data science methods and wind energy applications Includes specific demonstration of particular data science methods and their use in the context of addressing wind energy needs Presents real data, case studies and computer codes from wind energy research and industrial practice Covers material based on the author's ten plus years of academic research and insights

Table of Contents

Cover -- Half Title -- Title Page -- Copyright Page -- Dedication -- Table of Contents -- Foreword -- Preface -- Acknowledgments -- Chapter 1: Introduction -- 1.1 WIND ENERGY BACKGROUND -- 1.2 ORGANIZATION OF THIS BOOK -- 1.2.1 Who Should Use This Book -- 1.2.2 Note for Instructors -- 1.2.3 Datasets Used in the Book -- Part I: Wind Field Analysis -- Chapter 2: A Single Time Series Model -- 2.1 TIME SCALE IN SHORT- TERM FORECASTING -- 2.2 SIMPLE FORECASTING MODELS -- 2.2.1 Forecasting Based on Persistence Model -- 2.2.2 Weibull Distribution -- 2.2.3 Estimation of Parameters in Weibull Distribution -- 2.2.4 Goodness of Fit -- 2.2.5 Forecasting Based on Weibull Distribution -- 2.3 DATA TRANSFORMATION AND STANDARDIZATION -- 2.4 AUTOREGRESSIVE MOVING AVERAGE MODELS -- 2.4.1 Parameter Estimation -- 2.4.2 Decide Model Order -- 2.4.3 Model Diagnostics -- 2.4.4 Forecasting Based on ARMA Model -- 2.5 OTHER METHODS -- 2.5.1 Kalman Filter -- 2.5.2 Support Vector Machine -- 2.5.3 Artificial Neural Network -- 2.6 PERFORMANCE METRICS -- 2.7 COMPARING WIND FORECASTING METHODS -- Chapter 3: Spatio temporal Models -- 3.1 COVARIANCE FUNCTIONS AND KRIGING -- 3.1.1 Properties of Covariance Functions -- 3.1.2 Power Exponential Covariance Function -- 3.1.3 Kriging -- 3.2 SPATIO-TEMPORAL AUTOREGRESSIVE MODELS -- 3.2.1 Gaussian Spatio-temporal Autoregressive Model -- 3.2.2 Informative Neighborhood -- 3.2.3 Forecasting and Comparison -- 3.3 SPATIO-TEMPORAL ASYMMETRY AND SEPARABILITY -- 3.3.1 Definition and Quantification -- 3.3.2 Asymmetry of Local Wind Field -- 3.3.3 Asymmetry Quantification -- 3.3.4 Asymmetry and Wake Effect -- 3.4 ASYMMETRIC SPATIO-TEMPORAL MODELS -- 3.4.1 Asymmetric Non-separable Spatio-temporal Model -- 3.4.2 Separable Spatio-temporal Models -- 3.4.3 Forecasting Using Spatio-temporal Model -- 3.4.4 Hybrid of Asymmetric Model and SVM -- 3.5 CASE STUDY.

Chapter 4: Regime-switching Methods for Forecasting -- 4.1 REGIME-SWITCHING AUTOREGRESSIVE MODEL -- 4.1.1 Physically Motivated Regime Definition -- 4.1.2 Data-driven Regime Determination -- 4.1.3 Smooth Transition between Regimes -- 4.1.4 Markov Switching between Regimes -- 4.2 REGIME-SWITCHING SPACE-TIME MODEL -- 4.3 CALIBRATION IN REGIME SWITCHING METHOD -- 4.3.1 Observed Regime Changes -- 4.3.2 Unobserved Regime Changes -- 4.3.3 Framework of Calibrated Regime-switching -- 4.3.4 Implementation Procedure -- 4.4 CASE STUDY -- 4.4.1 Modeling Choices and Practical Considerations -- 4.4.2 Forecasting Results -- Part II: Wind Turbine Performance Analysis -- Chapter 5: Power Curve Modeling and Analysis -- 5.1 IEC BINNING: SINGLE-DIMENSIONAL POWER CURVE -- 5.2 KERNEL-BASED MULTI-DIMENSIONAL POWER CURVE -- 5.2.1 Need for Nonparametric Modeling Approach -- 5.2.2 Kernel Regression and Kernel Density Estimation -- 5.2.3 Additive Multiplicative Kernel Model -- 5.2.4 Bandwidth Selection -- 5.3 OTHER DATA SCIENCE METHODS -- 5.3.1 k-Nearest Neighborhood Regression -- 5.3.2 Tree-based Regression -- 5.3.3 Spline-based Regression -- 5.4 CASE STUDY -- 5.4.1 Model Parameter Estimation -- 5.4.2 Important Environmental Factors Affecting Power Output -- 5.4.3 Estimation Accuracy of Different Models -- Chapter 6: Production Efficiency Analysis and Power Curve -- 6.1 THREE EFFICIENCY METRICS -- 6.1.1 Availability -- 6.1.2 Power Generation Ratio -- 6.1.3 Power Coefficient -- 6.2 COMPARISON OF EFFICIENCY METRICS -- 6.2.1 Distributions -- 6.2.2 Pairwise Differences -- 6.2.3 Correlations and Linear Relationships -- 6.2.4 Overall Insight -- 6.3 A SHAPE-CONSTRAINED POWER CURVE MODEL -- 6.3.1 Background of Production Economics -- 6.3.2 Average Performance Curve -- 6.3.3 Production Frontier Function and Effi ciency Metric -- 6.4 CASE STUDY.

Chapter 7: Quantification of Turbine Upgrade -- 7.1 PASSIVE DEVICE INSTALLATION UPGRADE -- 7.2 COVARIATE MATCHING BASED APPROACH -- 7.2.1 Hierarchical Subgrouping -- 7.2.2 One-to-One Matching -- 7.2.3 Diagnostics -- 7.2.4 Paired t-tests and Upgrade Quantification -- 7.2.5 Sensitivity Analysis -- 7.3 POWER CURVE-BASED APPROACH -- 7.3.1 The Kernel Plus Method -- 7.3.2 Kernel Plus Quantification Procedure -- 7.3.3 Upgrade Detection -- 7.3.4 Upgrade Quantification -- 7.4 AN ACADEMIA-INDUSTRY CASE STUDY -- 7.4.1 The Power-vs-Power Method -- 7.4.2 Joint Case Study -- 7.4.3 Discussion -- 7.5 COMPLEXITIES IN UPGRADE QUANTIFICATION -- Chapter 8: Wake Effect Analysis -- 8.1 CHARACTERISTICS OF WAKE EFFECT -- 8.2 JENSEN'S MODEL -- 8.3 A DATA BINNING APPROACH -- 8.4 SPLINE-BASED SINGLE-WAKE MODEL -- 8.4.1 Baseline Power Production Model -- 8.4.2 Power Diff erence Model for Two Turbines -- 8.4.3 Spline Model with Non-negativity Constraint -- 8.5 GAUSSIAN MARKOV RANDOM FIELD MODEL -- 8.6 CASE STUDY -- 8.6.1 Performance Comparison of Wake Models -- 8.6.2 Analysis of Turbine Wake Effect -- Part III: Wind Turbine Reliability Management -- Chapter 9: Overview of Wind Turbine Maintenance Opti- mization -- 9.1 COST- EFFECTIVE MAINTENANCE -- 9.2 UNIQUE CHALLENGES IN TURBINE MAINTENANCE -- 9.3 COMMON PRACTICES -- 9.3.1 Failure Statistics-Based Approaches -- 9.3.2 Physical Load-Based Reliability Analysis -- 9.3.3 Condition-Based Monitoring or Maintenance -- 9.4 DYNAMIC TURBINE MAINTENANCE OPTIMIZATION -- 9.4.1 Partially Observable Markov Decision Process -- 9.4.2 Maintenance Optimization Solutions -- 9.4.3 Integration of Optimization and Simulation -- 9.5 DISCUSSION -- Chapter 10: Extreme Load Analysis -- 10.1 FORMULATION FOR EXTREME LOAD ANALYSIS -- 10.2 GENERALIZED EXTREME VALUE DISTRIBUTIONS -- 10.3 BINNING METHOD FOR NONSTATIONARY GEV DISTRIBUTION.

10.4 BAYESIAN SPLINE-BASED GEV MODEL -- 10.4.1 Conditional Load Model -- 10.4.2 Posterior Distribution of Parameters -- 10.4.3 Wind Characteristics Model -- 10.4.4 Posterior Predictive Distribution -- 10.5 ALGORITHMS USED IN BAYESIAN INFERENCE -- 10.6 CASE STUDY -- 10.6.1 Selection of Wind Speed Model -- 10.6.2 Pointwise Credible Intervals -- 10.6.3 Binning versus Spline Methods -- 10.6.4 Estimation of Extreme Load -- 10.6.5 Simulation of Extreme Load -- Chapter 11: Computer Simulator-Based Load Analysis -- 11.1 TURBINE LOAD COMPUTER SIMULATION -- 11.1.1 NREL Simulators -- 11.1.2 Deterministic and Stochastic Simulators -- 11.1.3 Simulator versus Emulator -- 11.2 IMPORTANCE SAMPLING -- 11.2.1 Random Sampling for Reliability Analysis -- 11.2.2 Importance Sampling Using Deterministic Simulator -- 11.3 IMPORTANCE SAMPLING USING STOCHASTIC SIMULATORS -- 11.3.1 Stochastic Importance Sampling Method 1 -- 11.3.2 Stochastic Importance Sampling Method 2 -- 11.3.3 Benchmark Importance Sampling Method -- 11.4 IMPLEMENTING STOCHASTIC IMPORTANCE SAMPLING -- 11.4.1 Modeling the Conditional POE -- 11.4.2 Sampling from Importance Sampling Densities -- 11.4.3 The Algorithm -- 11.5 CASE STUDY -- 11.5.1 Numerical Analysis -- 11.5.2 NREL Simulator Analysis -- Chapter 12: Anomaly Detection and Fault Diagnosis -- 12.1 BASICS OF ANOMALY DETECTION -- 12.1.1 Types of Anomalies -- 12.1.2 Categories of Anomaly Detection Approaches -- 12.1.3 Performance Metrics and Decision Process -- 12.2 BASICS OF FAULT DIAGNOSIS -- 12.2.1 Tree-Based Diagnosis -- 12.2.2 Signature-Based Diagnosis -- 12.3 SIMILARITY METRICS -- 12.3.1 Norm and Distance Metrics -- 12.3.2 Inner Product and Angle-Based Metrics -- 12.3.3 Statistical Distance -- 12.3.4 Geodesic Distance -- 12.4 DISTANCE-BASED METHODS -- 12.4.1 Nearest Neighborhood-based Method -- 12.4.2 Local Outlier Factor.

12.4.3 Connectivity-based Outlier Factor -- 12.4.4 Subspace Outlying Degree -- 12.5 GEODESIC DISTANCE BASED METHOD -- 12.5.1 Graph Model of Data -- 12.5.2 MST Score -- 12.5.3 Determine Neighborhood Size -- 12.6 CASE STUDY -- 12.6.1 Benchmark Cases -- 12.6.2 Hydropower Plant Case -- Bibliography -- Index.