Neural Network based Model Reference Control for Electric Heating Furnace with Input Saturation

Investigating gas furnace control practices with reinforcement learning

“…The quantity being optimized is the distribution of the inflow between the 6 injectors V ioe[ [1,6]] .…”

“…Temperature control has been addressed by a lot of articles in the literature, and with the emergence of Artificial Neural Networks (ANNs) taking advantage of the most recent advances in computational power and data analysis, new control techniques that are more robust and can deal with non linearity, inertia and time delay have been developed : image processing and clustering to control gas burners [4,5], recurrent neural networks (RNNs) with electrical furnaces [6,7], neural networks for metal quality control [8] and for flame stability control [9][10][11], and radial basis function neural networks (RBFNNs) to control coke furnaces [12]. In the aerospace industry, burners are also very thoroughly studied to prevent any incident inside aircraft engines.…”