Advanced Ensemble Model for Solar Radiation Forecasting Using Sine Cosine Algorithm and Newton’s Laws

As research in alternate energy sources is growing, solar radiation catching the eyes of community immensely. Since generation depends on uncontrollable natural variables, without proper forecasting, this source cannot be trusted. For use machine learning algorithms one best choices. This paper proposed an optimized forecasting ensemble model consisting pre-processing and training phases. The phase works advanced sine cosine algorithm (ASCA) using Newton’s laws gravity motion for objects (agents). ASCA uses functions to update agent’s position/velocity components by considering its mass. then developed k-nearest neighbors (KNN) regression. performance measured a dataset from Kaggle (Solar Radiation Prediction, Task NASA Hackathon). evaluated comparison with Particle Swarm Optimizer (PSO), Whale Optimization Algorithm (WOA), Genetic (GA), Grey Wolf (GWO), Squirrel Search (SSA), Harris Hawks (HHO), Hybrid Greedy Sine Cosine Differential Evolution (HGSCADE), Modified Cuckoo (HMSCACSA), Marine Predators (MPA), Chimp (ChOA), Slime Mould (SMA). Obtained results are compared those state-of-the-art models, significant superiority confirmed statistical analysis such as ANOVA Wilcoxon’s rank-sum tests.