Stack effect is a dominant driving force for building natural ventilation.Analytical models were developed for the evaluation of stack effect in a shaft,accounting for the heat transfer from shaft interior boundaries....Stack effect is a dominant driving force for building natural ventilation.Analytical models were developed for the evaluation of stack effect in a shaft,accounting for the heat transfer from shaft interior boundaries.Both the conditions with constant heat flux from boundaries to the airflow and the ones with constant boundary temperature were considered.The prediction capabilities of these analytical models were evaluated by using large eddy simulation(LES) for a hypothetical shaft.The results show that there are fairly good agreements between the predictions of the analytical models and the LES predictions in mass flow rate,vertical temperatures profile and pressure difference as well.Both the results of analytical models and LES show that the neutral plane could locate higher than one half of the shaft height when the upper opening area is identical with the lower opening area.Further,it is also shown that the analytical models perform better than KLOTE's model does in the mass flow rate prediction.展开更多
Experimentation data of perspex glass sheet cutting, using CO2 laser, with missing values were modelled with semi-supervised artificial neural networks. Factorial design of experiment was selected for the verification...Experimentation data of perspex glass sheet cutting, using CO2 laser, with missing values were modelled with semi-supervised artificial neural networks. Factorial design of experiment was selected for the verification of orthogonal array based model prediction. It shows improvement in modelling of edge quality and kerf width by applying semi-supervised learning algorithm, based on novel error assessment on simulations. The results are expected to depict better prediction on average by utilizing the systematic randomized techniques to initialize the neural network weights and increase the number of initialization. Missing values handling is difficult with statistical tools and supervised learning techniques; on the other hand, semi-supervised learning generates better results with the smallest datasets even with missing values.展开更多
基金Project(50838009) supported by the National Natural Science Foundation of ChinaProject(2010DFA72740-03) supported by the National Key Technology Research and Development Program of China
文摘Stack effect is a dominant driving force for building natural ventilation.Analytical models were developed for the evaluation of stack effect in a shaft,accounting for the heat transfer from shaft interior boundaries.Both the conditions with constant heat flux from boundaries to the airflow and the ones with constant boundary temperature were considered.The prediction capabilities of these analytical models were evaluated by using large eddy simulation(LES) for a hypothetical shaft.The results show that there are fairly good agreements between the predictions of the analytical models and the LES predictions in mass flow rate,vertical temperatures profile and pressure difference as well.Both the results of analytical models and LES show that the neutral plane could locate higher than one half of the shaft height when the upper opening area is identical with the lower opening area.Further,it is also shown that the analytical models perform better than KLOTE's model does in the mass flow rate prediction.
文摘Experimentation data of perspex glass sheet cutting, using CO2 laser, with missing values were modelled with semi-supervised artificial neural networks. Factorial design of experiment was selected for the verification of orthogonal array based model prediction. It shows improvement in modelling of edge quality and kerf width by applying semi-supervised learning algorithm, based on novel error assessment on simulations. The results are expected to depict better prediction on average by utilizing the systematic randomized techniques to initialize the neural network weights and increase the number of initialization. Missing values handling is difficult with statistical tools and supervised learning techniques; on the other hand, semi-supervised learning generates better results with the smallest datasets even with missing values.
