Eine Plattform für die Wissenschaft: Bauingenieurwesen, Architektur und Urbanistik
Prediction method of coincident design day for design cooling load calculation
Highlights A decision chain method for predicting coincident design day (CDD) proposed. Correlation analysis and Support Vector Machine used to construct decision chain. Performance of decision chain improved through combination of local optimizations. Predicted CDD meets the requirements of practical applications in almost cases.
Abstract Coincident design day (CDD) means the design day that considered the simultaneous occurrence of design weather elements and the correlation between design weather data and room parameters. The application of coincident design day will make the load calculation more accurate in the air-conditioning system design. However, since there are excessive possible combinations of the room parameters in practical engineering applications, it is an urgent problem to predict or match the CDD of designed room from existing/typical CDD set. In this study, a CDD prediction method based on Support Vector Machine and decision chain is proposed to solve this problem. The practicality of the prediction method was verified by evaluating its performance. The test results for Hong Kong show that average 81.71 % cases are within ±1.0 % of the deviation between the design cooling load calculated by predicted CDD and actual design cooling load, 98.02 % cases are in within ±3.0 %, and 99.33 % cases are within ±5.0 %. The verifying results of Changsha show that average 82.16 % cases are within ±1.0 % of the deviation, 97.58 % cases are within ±3.0 %, and 99.53 % cases are within ±5.0 %. These indicate that the decision chain method is practical in engineering applications. This study filled the gap of the CDD application, which may provide a basis and some inspiration for subsequent studies on CDD.
Prediction method of coincident design day for design cooling load calculation
Highlights A decision chain method for predicting coincident design day (CDD) proposed. Correlation analysis and Support Vector Machine used to construct decision chain. Performance of decision chain improved through combination of local optimizations. Predicted CDD meets the requirements of practical applications in almost cases.
Abstract Coincident design day (CDD) means the design day that considered the simultaneous occurrence of design weather elements and the correlation between design weather data and room parameters. The application of coincident design day will make the load calculation more accurate in the air-conditioning system design. However, since there are excessive possible combinations of the room parameters in practical engineering applications, it is an urgent problem to predict or match the CDD of designed room from existing/typical CDD set. In this study, a CDD prediction method based on Support Vector Machine and decision chain is proposed to solve this problem. The practicality of the prediction method was verified by evaluating its performance. The test results for Hong Kong show that average 81.71 % cases are within ±1.0 % of the deviation between the design cooling load calculated by predicted CDD and actual design cooling load, 98.02 % cases are in within ±3.0 %, and 99.33 % cases are within ±5.0 %. The verifying results of Changsha show that average 82.16 % cases are within ±1.0 % of the deviation, 97.58 % cases are within ±3.0 %, and 99.53 % cases are within ±5.0 %. These indicate that the decision chain method is practical in engineering applications. This study filled the gap of the CDD application, which may provide a basis and some inspiration for subsequent studies on CDD.
Prediction method of coincident design day for design cooling load calculation
Fang, Zhengcheng (Autor:in) / Chen, Youming (Autor:in) / Ai, Zhengtao (Autor:in) / Li, Hongqiang (Autor:in)
Energy and Buildings ; 276
20.09.2022
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
New design-day method for building cooling load calculation in China
Taylor & Francis Verlag | 2019
|An Occupant-Oriented Calculation Method of Building Interior Cooling Load Design
DOAJ | 2018
|