Policies for complex visual tasks have been successfully learned with deep reinforcement learning, using an approach called deep Q-networks (DQN), but relatively large (task-specific) networks and extensive training are needed to achieve good performance. In this work, we present a novel method called policy distillation that can be used to extract the policy of a reinforcement learning agent a...

Further Reading Geankoplis CJ (1993) Transport Processes and Unit Operations. New Jersey: PTR Prentice Hall. King CJ (1980) Separation Processes. New York:McGrawHill. Kister HZ (1992) Distillation Design. New York: McGraw-Hill. Seader JD and Henley EJ (1998) Separation Process Principles. New York: John Wiley. Stichlmair JG and Fair JR (1998)Distillation Principles and Practice. New York: Wiley...

The problem of separating the liquid components which have been recovered from natural gas is discussed. Opportunities for improved efficiency using distributed distillation, recycle coupling, heat pumping, and partial effect distillation are presented. Emphasis is placed upon energy efficiency with potential savings in capital costs as applied to heat driven processes.

Huiping Song, Xingang LI, Jinsheng Sun and Yanhong Wang, (1)School of Chemical Engineering and Technology, Tianjin University, No.92 Weijin Road, Nankai district, Tianjin, 300072, China, (2)School of Chemical Engineering and Technology, National Engineering Research Center for Distillation Technology, Tianjin University,P.R. China, Tianjin, China, (3)National Engineering Research Center for Dis...

This paper clearly shows that the recently proposed terrain methodology of Lucia and Yang (2003) can be used to solve steady-state distillation problems in a reliable and efficient manner. Numerical results are presented that show that terrain methods are superior to Newton’s method and homotopy-continuation for distillation examples with multiple solutions. © 2004 Elsevier Ltd. All rights rese...

Batch industries, specially the fine chemical industry, need rapid and robust mathematical models that can easily predict the degree of separation of mixtures like volatile organic compounds. Otherwise, in the study of on line production scheduling and planning of fine chemical plants, there is a lack of quick but reliable and more accurate than black box models in batch distillation. In this w...

Problem statement: We have recently introduced several important improvements in the measurement of distillation curves for complex fluids. The modifications include a compositionexplicit data channel for each distillate fraction and temperature measurements that are true thermodynamic state points that can be modeled with an equation of state. The composition-explicit information is achieved w...