SearchSkip to Search Results
- 2Reinforcement Learning
- 1Artificial Neural Network
- 1Design of experiments
- 1Experimental data
- 1First Principle Modeling
Real-time optimization systems have become a common tool, in the continuous manufacturing industries, for improving process performance. Typically, these are on-line, steady-state, model-based optimization systems, whose eﬀectiveness depends on a large number of design decisions. The work...
The Hybrid tank pilot plant was designed in Process Control Laboratory (PCL) of the University of Alberta at 2009. In this report, we try to construct a model for this process. Using the physical behavior of the plant, it is possible to have the nonlinear first principle dynamic model of that...
Poor control of hot strip mill loopers degrades strip width and gauge, and may even lead to mill breakdowns due to instability. In this study, dynamics of the looper-strip system and the control challenges it poses are discussed, and covariance control theory is applied to variance control design...
A Long-Short Term Memory Recurrent Neural Network Based Reinforcement Learning Controller for Office Heating Ventilation and Air Conditioning SystemsDownload
Energy optimization in buildings by controlling the Heating Ventilation and Air Conditioning (HVAC) system is being researched extensively. In this paper, a model-free actor-critic Reinforcement Learning (RL) controller is designed using a variant of artificial recurrent neural networks called...
A Hierarchical Constrained Reinforcement Learning for Optimization of Bitumen Recovery Rate in a Primary Separation VesselDownload
This work proposes a two-level hierarchical constrained control structure for reinforcement learning (RL) with application in a Primary Separation Vessel (PSV). The lower level is concerned with servo tracking and regulation of the interface level against variances in ore quality by manipulating...