Solar Energy Forecasting and Resource Assessment

Chapter 15

Stochastic-Learning Methods

Carlos F.M. Coimbra and Hugo T.C. Pedro,    Center of Excellence in Renewable Energy Integration, Department of Mechanical and Aerospace Engineering,    Jacobs School of Engineering, University of California, San Diego

Chapter Outline

15.1. Introduction

15.2. Baseline Methods for Comparison

15.2.1. Persistence Methods

15.2.2. ARIMA Models

15.2.3. kNN and ANN

15.3. Genetic Algorithms

15.3.1. GA/ANN: Scanning the Solution Space

15.3.2. Selection, Crossover, Mutation, and Stopping Criterion

15.4. Qualitative Performance Assessment

15.5. Performance of Stochastic-Learning Methods with No Exogenous Variables

15.5.1. Clear-Sky Model

15.5.2. Quantitative Performance of ARIMA, kNN, ANN, GA/ANN

15.6. Sky-Imaging Data ...