Bibliographic citations
Estrada, L., (2022). Diseño de un sistema de control basado en aprendizaje reforzado aplicado a un reactor industrial tipo Klatt-Engell [Tesis, Universidad Nacional de Ingeniería]. http://hdl.handle.net/20.500.14076/22804
Estrada, L., Diseño de un sistema de control basado en aprendizaje reforzado aplicado a un reactor industrial tipo Klatt-Engell [Tesis]. PE: Universidad Nacional de Ingeniería; 2022. http://hdl.handle.net/20.500.14076/22804
@mastersthesis{renati/712166,
title = "Diseño de un sistema de control basado en aprendizaje reforzado aplicado a un reactor industrial tipo Klatt-Engell",
author = "Estrada Rayme, Leighton Leandro",
publisher = "Universidad Nacional de Ingeniería",
year = "2022"
}
In this research, a control system based in reinforcement learning approach is developed to control the temperature and product concentration of an industrial reactor Klatt-Engell represented by a mathematical model of non-linear behavior. The design of the control system follows a methodology based on the improvement of the model free learning control scheme. It includes elements of reinforcement learning such as the use of the Q-learning algorithm, learning parameters, exploration and exploitation Parameters. The representation of the control system consists of a block diagram, with two model free learning control subsystems that control each variable. In addition, before executing the control of the plant, the calculation of the optimal values necessary for the good choice of the control signals is required through a training process. Finally, the results obtained for different reference signals, exogenous input signals and presence of noise in the output variables are shown, and the comparison of these results by means of a performance index against other control systems.
This item is licensed under a Creative Commons License