| Heat storage plays an important role in social energy conservation and sustainable economic development.The heat storage performance of wooden flooring for heating floor reflects the thermal stability and thermal storage ability of wooden flooring,and it is one of the important indexes to characterize the thermal comfort of wooden flooring.Because the wooden flooring is an anisotropic heterogeneous material,it is difficult to evaluate its thermal storage performance according to the existing methods and theories describing the thermal properties of homogeneous materials.It is necessary to study the technology and evaluation system for testing the thermal storage performance of heterogeneous materials.Therefore,based on the heat storage performance of wooden flooring,the temperature field prediction model of fuzzy and support vector machine is established,and the thermal storage performance of wooden flooring is studied by this model.The purpose of this paper is to provide guidance for the development of testing technology and evaluation standards for thermal properties of wooden flooring used in floor heating industry.Firstly,in order to verify the feasibility of using support vector machine to model the temperature field,according to the testing instrument of floor thermal storage performance developed by the team,the physical model of cylindrical closed cavity is constructed in this paper.The mass,momentum and energy conservation equations of heat field heat transfer are established by the law of fluid mechanics,and solved by computational fluid dynamics software(CFD),and the characteristic data of temperature field in the closed cavity are obtained.The classical support vector machine(SVM)model is trained and verified with this data,and a good result is obtained.Secondly,based on the temperature field data collected by ground heating floor thermal storage performance testing instrument,three models of fuzzy classical support vector machine(Fuzzy & SVM),fuzzy least squares support vector machine(Fuzzy & LSSVM)and fuzzy twin support vector machine(Fuzzy & TSVM)are established.In the process of modeling,the temperature field data collected are divided into four fuzzy subsets according to the time dimension,and the support vector machine model is built in each fuzzy sub-set.Then the trapezoidal fuzzy function is constructed with the time as the variable.The fuzzy function is used to assign different membership weights to the predicted results of each model and make fuzzy superposition.The output of the superposition is taken as the final result of the whole model.The modeling results show that the fuzzy support vector machine model can accurately describe the temperature field in the closed cavity.At the same time,the introduction of fuzzy method effectively solves the problem of "dimension disaster" caused by matrix inversion,and makes up for the limitation of support vector machine in modeling large sample data.Finally,in order to accurately define the thermal storage performance of wooden floorings,this concept is associated with the heat released by wooden floorings in the closed chamber,the volume of the floor and the time of temperature balance,thus improving the scientific nature of the definition.When calculating the heat absorbed by the air in the closed chamber,the idea of space element is put forward.The closed chamber is divided into several small space microelements first,and then the heat in each space micro-element is calculated by the temperature value obtained by the support vector machine model,and the heat in each space micro-element is calculated by the temperature obtained from the support vector machine model.This method can effectively improve the detection accuracy of heat storage performance.According to the above-mentioned methods,the thermal storage properties of 10 different wooden floorings of Betula platyphylla Suk.,Pinus koraiensis,Juglans mandshurica,Manglietia hainanensis,Ulmus rubra,Xylosma racemosum,Fraxinus mandshurica Rupr.,Betula alnoides,Schima wallichii,Homalium hainanense Gagnep.were tested.The equivalent specific heat capacity of the abovementioned materials is inversed by using the thermal storage performance,and the inversion results are compared with the experimental results in the literature.The results show that the error between the average value of the inversion specific heat capacity and the average specific heat capacity of the literature is within ±4.89%.This indicates that the experimental results obtained in this study are in good agreement with the experimental results in the literature,thus indirectly proving the accuracy of the fuzzy and support vector machine models used in this paper for the research results of the thermal storage performance of wooden floorings.The results of this study can provide a method and reference for the establishment of heat storage performance standards for wooden flooring. |
