A study of the method of the thermal conductivity measurement for VIPs with improved RBF neural networks. (June 2016)

• Record Type: Journal Article
• Authors: Xia, Rongfei; Huang, Jun; Chen, Yifei; Feng, Yongjian
• Is Part Of: Measurement. Volume 87(2016:Jun.)
• Page Start: 246
• Page End: 254
• Publication Date: 2016-06
• Subjects: Vacuum insulation panel -- Thermal conductivity -- ANSYS -- RBF neural network -- Particle swarm optimization; Weights and measures -- Periodicals; Measurement -- Periodicals; Measurement; Weights and measures; Periodicals; 530.8
• DOI: 10.1016/j.measurement.2016.03.029
• Languages: English
• ISSNs: 0263-2241

Abstract

Highlights: A fast measuring method of VIP's thermal conductivity is proposed. The feasibility of the presented methods is proved by ANSYS simulation. The improved RBF neural network is presented to solve the nonlinear problems. Improved method overcomes many defects compared with the traditional method. Abstract: Thermal conductivity is an important physical quantity which is the token of the heat-conducting property of VIPs. The measurement of thermal conductivity is invaluable for VIPs can be widely used. This paper presented a fast and high precision measuring method of VIP's thermal conductivity, and proved its feasibility by simulation. New methods use improved neutral networks to fit the relationship between VIP's thermal conductivity and characteristic value of frequency change, and verified the superiority of this method by experiment compared with the least squares. The presented methods are very valuable for the widespread use of vacuum insulation panels.