Investigating Enhanced Cooling Load Estimation through Hybrid LSSVR Models

Rising global urbanization necessitates accurate energy consumption prediction, particularly for residential buildings. Given the significant influence of cooling systems on operational costs, it is valuable to explore models for forecasting building heating and cooling loads. Unlike many prior stud...

Full description

Saved in:
Bibliographic Details
Main Authors: Ali Hassan, Hamza Rashid
Format: Article
Language:English
Published: Bilijipub publisher 2024-03-01
Series:Journal of Artificial Intelligence and System Modelling
Subjects:
Online Access:https://jaism.bilijipub.com/article_193315_58803f7bda3f23fdeed6f5df0adb899c.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849389551053701120
author Ali Hassan
Hamza Rashid
author_facet Ali Hassan
Hamza Rashid
author_sort Ali Hassan
collection DOAJ
description Rising global urbanization necessitates accurate energy consumption prediction, particularly for residential buildings. Given the significant influence of cooling systems on operational costs, it is valuable to explore models for forecasting building heating and cooling loads. Unlike many prior studies relying on historical energy data, this research emphasizes incorporating building attributes, including area and floor height, in developing predictive models for cooling loads. Cooling load prediction technology is pivotal to boosting energy efficiency in cooling systems. Conventional computational models, encompassing both forward and inverse modeling, demand substantial computational resources and prolonged procedures. Conversely, artificial intelligence exhibits superior performance, employing adaptable models adept at pattern recognition and self-improvement as data accumulates. These models excel at forecasting cooling loads, incorporating weather conditions, building materials, and occupancy, yielding swift, responsive predictions that augment energy efficiency. The utilized dataset contains 768 samples, which are extracted from published articles. This research examines the application of Least Squares Support Vector Regression (LS-SVR) in predicting cooling load consumption for buildings. The LS-SVR model is enhanced through optimization techniques such as Sea Horse Optimizer (SHO) and Runge Kutta Optimization (RKO). The results reveal that the LSSVR+SHO (LSSH) prediction model surpasses the LSSVR in cooling load predictions, as reflected in its Root Mean Square Error (RMSE) of 1.019, which is approximately 0.666 lower than the LSSVR. Additionally, the LSSH model exhibits exceptional performance, particularly during the training phase, achieving an optimal R2 value of 0.988.
format Article
id doaj-art-e5e274972af14745a89cd611d3cb3768
institution Kabale University
issn 3041-850X
language English
publishDate 2024-03-01
publisher Bilijipub publisher
record_format Article
series Journal of Artificial Intelligence and System Modelling
spelling doaj-art-e5e274972af14745a89cd611d3cb37682025-08-20T03:41:56ZengBilijipub publisherJournal of Artificial Intelligence and System Modelling3041-850X2024-03-010201152810.22034/jaism.2024.443912.1024193315Investigating Enhanced Cooling Load Estimation through Hybrid LSSVR ModelsAli Hassan0Hamza Rashid1Department of Mechatronics Engineering, Center of Industrial Electronics (CIE), University of Southern Denmark, Sønderborg, 6400, DenmarkDepartment of Mechatronics Engineering, Center of Industrial Electronics (CIE), University of Southern Denmark, Sønderborg, 6400, DenmarkRising global urbanization necessitates accurate energy consumption prediction, particularly for residential buildings. Given the significant influence of cooling systems on operational costs, it is valuable to explore models for forecasting building heating and cooling loads. Unlike many prior studies relying on historical energy data, this research emphasizes incorporating building attributes, including area and floor height, in developing predictive models for cooling loads. Cooling load prediction technology is pivotal to boosting energy efficiency in cooling systems. Conventional computational models, encompassing both forward and inverse modeling, demand substantial computational resources and prolonged procedures. Conversely, artificial intelligence exhibits superior performance, employing adaptable models adept at pattern recognition and self-improvement as data accumulates. These models excel at forecasting cooling loads, incorporating weather conditions, building materials, and occupancy, yielding swift, responsive predictions that augment energy efficiency. The utilized dataset contains 768 samples, which are extracted from published articles. This research examines the application of Least Squares Support Vector Regression (LS-SVR) in predicting cooling load consumption for buildings. The LS-SVR model is enhanced through optimization techniques such as Sea Horse Optimizer (SHO) and Runge Kutta Optimization (RKO). The results reveal that the LSSVR+SHO (LSSH) prediction model surpasses the LSSVR in cooling load predictions, as reflected in its Root Mean Square Error (RMSE) of 1.019, which is approximately 0.666 lower than the LSSVR. Additionally, the LSSH model exhibits exceptional performance, particularly during the training phase, achieving an optimal R2 value of 0.988.https://jaism.bilijipub.com/article_193315_58803f7bda3f23fdeed6f5df0adb899c.pdfbuilding cooling load predictionprediction modelsleast square support vector regressionsea horse optimizerrunge kutta optimization
spellingShingle Ali Hassan
Hamza Rashid
Investigating Enhanced Cooling Load Estimation through Hybrid LSSVR Models
Journal of Artificial Intelligence and System Modelling
building cooling load prediction
prediction models
least square support vector regression
sea horse optimizer
runge kutta optimization
title Investigating Enhanced Cooling Load Estimation through Hybrid LSSVR Models
title_full Investigating Enhanced Cooling Load Estimation through Hybrid LSSVR Models
title_fullStr Investigating Enhanced Cooling Load Estimation through Hybrid LSSVR Models
title_full_unstemmed Investigating Enhanced Cooling Load Estimation through Hybrid LSSVR Models
title_short Investigating Enhanced Cooling Load Estimation through Hybrid LSSVR Models
title_sort investigating enhanced cooling load estimation through hybrid lssvr models
topic building cooling load prediction
prediction models
least square support vector regression
sea horse optimizer
runge kutta optimization
url https://jaism.bilijipub.com/article_193315_58803f7bda3f23fdeed6f5df0adb899c.pdf
work_keys_str_mv AT alihassan investigatingenhancedcoolingloadestimationthroughhybridlssvrmodels
AT hamzarashid investigatingenhancedcoolingloadestimationthroughhybridlssvrmodels