Machine-learning-physics-based-hybrid-modeling

A Quick Wrap-up of My Second Undergraduate Research (2022.6-2022.8)

Our research focuses on using machine learning techniques to evaluate and simulate cooling systems. In this project, I collaborated with industrial partners to utilize real-world cooling system operation data and develop a hybrid energy model using Python. This research emphasized calibrating the hybrid energy model with advanced machine learning techniques such as Xgboost, LSTM, and Random Forest algorithms to enhance predictive accuracy. The research successfully constructed a high-performance model, achieving a 78% accuracy rate, demonstrating its potential for real-world applications in the energy sector. This repository includes all machine learning methods used to simulate the cooling system model, enabling the company to predict the operating conditions for upcoming years and make relevant adjustments to achieve high performance and enhance energy efficiency, ultimately contributing to a more environmentally friendly approach.