https://github.com/mdarshad1000/hvac-predictive-maintenance-paper

https://github.com/mdarshad1000/hvac-predictive-maintenance-paper

Last synced: about 6 hours ago
JSON representation

• Host: GitHub
• URL: https://github.com/mdarshad1000/hvac-predictive-maintenance-paper
• Owner: mdarshad1000
• Created: 2023-11-25T15:51:43.000Z (12 months ago)
• Default Branch: master
• Last Pushed: 2023-11-25T16:05:25.000Z (12 months ago)
• Last Synced: 2024-05-01T14:00:58.478Z (7 months ago)
• Language: Jupyter Notebook
• Size: 5.65 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Predictive Maintenance Strategies for HVAC Systems: Leveraging MPC, Dynamic Energy Performance Analysis, and ML Classification Models

This research explores the multifaceted challenges that can affect ventilation, air conditioning systems, and heating appliances, leading to reduced operational efficiency, heightened energy consumption, and increased maintenance expenses. Predictive maintenance, a progressive approach, is investigated as a pivotal strategy, complementing traditional HVAC equipment maintenance paradigms, including breakdown maintenance and preventative machine learning. Utilizing a diverse set of predictive models infused with machine learning techniques, this study employs the 'Semiconductor Manufacturing Process 

(SECOM) dataset' to simulate the manufacturing processes of HVAC systems, aligning with characteristics akin to semiconductor-based devices. The research undertakes a comparative analysis, contrasting the predictive capabilities of the Random Forest (RF) algorithm with the Support Vector Machine (SVM) in areas such as problem detection, diagnostics, and load monitoring. Notably, the RF model demonstrates superior prediction accuracy. The research aims to proactively detect potential HVAC system or component issues, discerning the nature of impending failures at their earliest stages to enable proactive maintenance strategies. Evaluation metrics such as the Receiver Operating Characteristic (ROC) curve and accuracy are employed for rigorous comparative analysis across various predictive machine learning classification models. Furthermore, a comprehensive 'dynamic energy performance benchmark' framework is meticulously developed for HVAC systems, facilitating real-time operational performance assessment and the identification of irregularities in power utilization at different operational stages. Additionally, Artificial Neural Network (ANN) models are employed to establish an administrative ModelPredictive Control (MPC) system tailored for residential HVAC applications.