cyrilvoyant/Make_Stationary: Make_Stationary - v1.0.0

This repository contains the MATLAB implementation of the PROJECTION and PHASE deseasonalization operators introduced in:

Agoundedemba et al., "Does Stationarization Help? A Frugal Benchmark of Deseasonalization for Renewable Energy Forecasting"

The code implements two frugal phase-conditioned Extreme Learning Machine (ELM) operators for removing deterministic cyclic components from renewable energy and meteorological time series (GHI, PV, wind speed, wind power, air temperature, total load). Both operators learn seasonal structure from normalized phase encodings and a single annual cycle, and estimate the seasonal component in closed form — without harmonic specification, multi-year records, or iterative tuning.