Knowledge distillation between machine learning models has opened many new avenues for parameter count reduction, performance improvements, or amortizing training time when changing architectures the teacher and student network. In case of reinforcement learning, this technique also been applied to distill policies students. Until now, policy required access a simulator real world trajectories....

Major pollutants emission from Direct-Injection (D.I) diesel engines are predicted by means of a mathematical model. In order to construct such a model, an integral Multi-Zone Combustion Model (MZCM) is formulated, which basically consists of jet mixing and combustion submodels. In MZCM, variation of injection pressure is considered. Also time period of ignition delay is predicted by considerin...

A long landing is one type of flight incident that will multiply the risk of a runway excursion. It occurs frequently but receives little attention in research due to difficulty in obtaining the real flight data. The aim of this paper is to discover key flight parameter features of long landing incidents by analyzing Quick Access Recorder (QAR) data and put forward prevention measures from the ...

For fundamental analysis of the transport of mass or heat in a dispersed system, accurate knowledge of behavior of jet is prerequisite. For generation of high interfacial area breaking of liquid jet in to small droplet is necessary. Liquid jet breakup is a complex phenomenon and dependent on many factors like velocity of jet, pressure difference, geometry of nozzle, temperature of both fluid, p...

[1] The distribution of daily North Atlantic jet latitude is analyzed in 45 CMIP3 integrations. It is demonstrated that models that place the jet equatorward of its observed position have more positively skewed jet lati tude distributions, while models that correctly place the jet have symmetric distributions like that of the observations. The jet is shown to be more persistent at equatorward l...