Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ Publikationer från K...arrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
Publikationer från KTH
Bachelor thesis . 2024
addClaim

This Research product is the result of merged Research products in OpenAIRE.

You have already added 0 works in your ORCID record related to the merged Research product.

Fault Diagnostics for Energy Systems through Machine Learning : A performance review of the implementation of ML methods for air conditioning, refrigeration and heat pump systems

Authors: Rengifo Palacios, Andres Felipe;

Fault Diagnostics for Energy Systems through Machine Learning : A performance review of the implementation of ML methods for air conditioning, refrigeration and heat pump systems

Abstract

Air conditioning, refrigeration and heat pumps use around 20%of theglobal electricity and it is estimated 25% of this is wasted due to poor commissioning, optimization, and maintenance. To minimize these problems and wastedenergy, predictive maintenance is a vital key to solve this issue. Due to the nature of thermal systems, its operation conditions are goingtovarysignificantly during different seasons for their components, thus comparingabaseline within models can be a subjective task if done through traditional COPmeasuring methods. Utilizing machine learning methods such as Multi-Layer Perceptrons andPCAanalysis, paired with ClimaCheck‘s internal method with the measurements of HVACsystems can help stablish these required baselines, and it can providespecificmeasurements of the efficiency expected of each component within theloop, providing warnings for decreases in performance caused by poor installationordegraded components. The method used led to highly reliable and fast results to determine virtual modelsfor the internal efficiencies of both condensers and evaporators with a lowRMSEof0.12% and coefficients R2 and D2 above 90% of correspondence with anaveragerunning time of 3 seconds for every hour of data. The performance does extendtoother values like the general internal efficiency (SEI) but does not extendwiththesame performance to KPIs based on temperature. Luftkonditionering, kylning och värmepumpar förbrukar cirka 20 % av den globalaelenergin, och det uppskattas att 25 % av denna energi går till spillo på grundavdåligdriftsättning, optimering och underhåll. För att minimera dessa problemochslöseri med energi är förebyggande underhåll en viktig nyckel för att lösa problemet. På grund av de termiska systemens natur kommer driftförhållandena att varieraavsevärt under olika årstider för deras komponenter, vilket innebär att det kanvaraensubjektiv uppgift att jämföra en baslinje inom modeller om det görs med traditionellaCOP-mätmetoder. Användning av maskininlärningsmetoder som Multi-Layer Perceptrons och PCA- analys, i kombination med ClimaChecks interna metod med mätningar av HVAC- system, kan hjälpa till att fastställa dessa nödvändiga baslinjer, och det kangespecifika mätningar av den effektivitet som förväntas av varje komponent i slingan, vilket ger varningar för minskad prestanda orsakad av dålig installation eller försämrade komponenter. Den metod som användes ledde till mycket tillförlitliga och snabba resultat för att fastställa virtuella modeller för den interna effektiviteten hos både kondensorer ochförångare med ett lågt RMSE på 0,12% och koefficienter R2 och D2 över 90%avöverensstämmelsen med en genomsnittlig körtid på 3 sekunder för varje timmemeddata. Prestandan sträcker sig till andra värden som den allmänna interna effektiviteten (SEI) men sträcker sig inte med samma prestanda till KPI: er baseradepå temperatur.

Country
Sweden
Related Organizations
Keywords

Maskininlärning, PCA, Compressor, Predictive Maintenance, kompressor, Energy Engineering, Efficiency, förångare, HVAC, condenser, Prediktivt underhåll, Regression, kondensor, Machine Learning, Energiteknik, regression, Effektivitet, AFDD, Evaporator

  • BIP!
    Impact byBIP!
    citations
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
citations
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Green
Related to Research communities
Energy Research