ANFIS Based Reinforcement Learning Strategy for Control A Nonlinear Coupled Tanks System

다국어 초록 (Multilingual Abstract)

In this paper, a novel algorithm based machine learning technique for control nonlinear coupled tanks system is presented.
An intelligent controller using adaptive neuro-fuzzy inference system (ANFIS) based reinforcement learning is proposed (ANFIS-RL) by representing the nonlinear coupled tanks system as a Markov decision process. A model-free learning algorithm has been used to train a policy that controls the liquid level of the tanks system without the need to determine the dynamic model of the controlled system. Based on the optimal learned policy, which is approximated by ANFIS, the controlled system can perform the best action quickly based on the states of the system. Simulation results demonstrated the feasibility of the proposed algorithm.

참고문헌 (Reference)

1 Haznedar B, "Training ANFIS structure using simulated annealing algorithm for dynamic systems identification" 302 : 66-74, 2018

2 García J, "Teaching a humanoid robot to walk faster through Safe Reinforcement Learning" 88 : 103360-, 2020

3 Biswas PP, "Sliding mode control of quadruple tank process" 19 (19): 548-561, 2009

4 Kara T, "Robust trajectory tracking control of robotic manipulators based on model-free PID-SMC approach" 6 (6): 170-188, 2018

5 Mary AH, "Robust proportional control for trajectory tracking of a nonlinear robotic manipulator: LMI optimization approach" 41 (41): 5027-5036, 2016

6 Li F, "Robot skill acquisition in assembly process using deep reinforcement learning" 345 : 92-102, 2019

7 Wu YH, "Reinforcement learning in dual-arm trajectory planning for a free-floating space robot" 98 : 105657-, 2020

8 Pane YP, "Reinforcement learning based compensation methods for robot manipulators" 78 : 236-247, 2019

9 Shipman WJ, "Reinforcement learning and deep neural networks for PI controller tuning" 52 (52): 111-116, 2019

10 Xu J, "Recurrent neural network for solving model predictive control problem in application of four-tank benchmark" 19 (19): 172-178, 2016

11 Patel HR, "Passive fault tolerant control system using feed-forward neural network for two-tank interacting conical level control system against partial actuator failures and disturbances" 52 (52): 141-146, 2019

12 Gouta H, "Observerbased backstepping controller for a state-coupled two-tank system" 61 (61): 259-268, 2015

13 Jegatheesh A, "Novel fuzzy fractional order PID controller for non linear interacting coupled spherical tank system for level process" 72 : 102948-, 2020

14 Mrugalski M, "Neural network-based robust actuator fault diagnosis for a non-linear multi-tank system" 61 : 318-328, 2016

15 Johannes Dornheim ; Norbert Link ; Peter Gumbsch, "Model-free Adaptive Optimal Control of Episodic Fixed-horizon Manufacturing Processes Using Reinforcement Learning" 제어·로봇·시스템학회 18 (18): 1593-1604, 2020

16 Başçi A, "Implementation of an adaptive fuzzy compensator for coupled tank liquid level control system" 91 : 12-18, 2016

17 Hamoudi AK, "Design an Integral Sliding Mode Controller for a Nonlinear System" 13 (13): 138-147, 2017

18 Boubakir A, "A neuro-fuzzy-sliding mode controller using nonlinear sliding surface applied to the coupled tanks system" 6 (6): 72-80, 2009