Prediction of Thermal Conductivity of Pure Liquids and Mixtures using Neural Network.
スポンサーリンク
概要
- 論文の詳細を見る
A new predictive tool exploiting neural network for evaluating thermal conductivity of liquids and mixtures at ambient or saturated pressures, is proposed. It covers a wide range on molecular species including hydrocarbons, alcohols, water as well as inorganics, with values extending over the range of 40 to 700 mW/m/K. A three-layer forward network has been trained using experimental data to provide a preliminary set of weights which are progressively refined. This strategy has been adopted so to make it possible to automatically update the weights as new information becomes available. The predictions are significantly better than any correlation or physico-chemically based models.
- The Society of Chemical Engineers, Japanの論文
著者
-
Lu M.
Department Of Chemistry Lanzhou University
-
Kam E.
Department Of Chemical Engineering The University Of Leeds
-
MCGREAVY C.
Department of Chemical Engineering, The University of Leeds
-
Kam E.
Department of Chemical Engineering, The University of Leeds
関連論文
- Magnetic properties of pattern nanowire arrays and its application in perpendicular recording (マルチメディアストレージ 第6回アジア情報記録技術シンポジウム〔英文〕)
- Magnetic Properties of Pattern Nanowire Arrays and Its Application In Perpendicular Recording
- Magnetization processes in ferromagnetic cubes from initial saturation states (マルチメディアストレージ--第5回アジア情報記録技術シンポジウム〔英文〕)
- Magnetization Processes In Ferromagnetic Cubes From Initial Staturation States
- On The Single Domain Size of Ferromagnetic Cubic Particles by Micromagnetics Simulation (第4回アジア情報記録技術シンポジウム〔英文〕)
- On The Single Domain Size of Ferromagnetic Cubic Particles By Micromagnetics Simulation
- On The Mesh Size In Numerical Micromagnetic Modeling Of Fine Ferromagnetic Particles
- On The Mesh Size In Numerical Micromagnetic Modeling Of Fine Ferromagnetic Particles
- ニューラルネットワークによる純液体および混合物の熱伝導度の推算
- Prediction of Thermal Conductivity of Pure Liquids and Mixtures using Neural Network.