Imitation Reinforcement Learning-Based Remote Rotary Inverted Pendulum Control in OpenFlow Network

Kim, Ju-Bong; Lim, Hyun-Kyo; Kim, Chan-Myung; Kim, Minsuk; Hong, Yong-Geun; Han, Youn-Hee

doi:10.1109/access.2019.2905621

Cited by 27 publications

(24 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…13, the proposed MLC converges to optimum solution more quickly and the evolved solution is better than the ones obtained by GA and ACO evolutionary algorithms. Compared with the traditional MLCs using supervised, unsupervised and reinforcement learning [39][40], the fuzzy structure is optimally determined via ABC computation to achieve optimal intelligent control.…”

Section: Experimental Results Of Trajectory Trackingmentioning

confidence: 99%

Artificial Bee Colony Optimization Algorithm Incorporated With Fuzzy Theory for Real-Time Machine Learning Control of Articulated Robotic Manipulators

Huang

Chuang

2020

IEEE Access

View full text Add to dashboard Cite

This paper presents a real-time machine learning control (MLC) of articulated robotic manipulators using artificial bee colony optimization (ABC) algorithm incorporated with fuzzy theory. The modified ABC with dynamic weight is used to optimize the fuzzy structure and fractional order. The fractional parameters, fuzzy membership functions and rule base are determined by means of the ABC computation. This ABC-fuzzy hybrid learning algorithm is applied to real-time MLC of robotic manipulators by including fractional order proportional-integral-derivative (FOPID) control strategy. The MLC's control gain parameters are online tuned via the ABC-fuzzy optimization. With the kinematics analysis of a sixdegree-of-freedom (DOF) articulated arm via reverse coordinates approach, an ABC-fuzzy MLC is developed to achieve motion control. A real-time operating system (RTOS) on a microprocessor collaborates with the ABC-fuzzy MLC to meet critical timing constraint by considering the dynamics of actuators. Finally, the mechatronic design and experimental setup of a six-DOF articulated robotic manipulator are constructed. Experimental results and comparative works are provided to demonstrate the merit of the proposed methods. Compared with the conventional control schemes, the proposed ABC-fuzzy MLC has theoretical and practice significance in term of real-time capability, online parameter tuning, convergent behavior and hybrid MLC. The proposed MLC methodologies are applicable to designing real-time modern controllers in both industry and academia. INDEX TERMS Artificial bee colony optimization, fuzzy theory, machine learning control, robotic arm.

show abstract

Section: Experimental Results Of Trajectory Trackingmentioning

confidence: 99%

Artificial Bee Colony Optimization Algorithm Incorporated With Fuzzy Theory for Real-Time Machine Learning Control of Articulated Robotic Manipulators

Huang

Chuang

2020

IEEE Access

View full text Add to dashboard Cite

show abstract

“…2) The proposed approach can accelerate the overall learning process for control policy through simulation as well as in real devices. 3) Using evolved Gradient Sharing and Transfer Learning, the proposed scheme reduce the learning time than our previous works [15]- [17]. 4) In addition, the proposed scheme based on Federated Reinforcement Learning achieves the desired goal with a little learning in the presence of much noise.…”

Section: Introductionmentioning

confidence: 95%

“…We are motivated by previous research trends to apply federated multi-agent reinforcement learning to multiple real devices and improve learning performance. In our previous research [15]- [17], we have applied the Deep Q Network (DQN) to the RIP system and IoT devices in the Software-Defined Network environments for automatic and training control policies.…”

Section: Introductionmentioning

confidence: 99%

Federated Reinforcement Learning Acceleration Method for Precise Control of Multiple Devices

et al. 2021

Self Cite

View full text Add to dashboard Cite

Nowadays, Reinforcement Learning (RL) is applied to various real-world tasks and attracts much attention in the fields of games, robotics, and autonomous driving. It is very challenging and devices overwhelming to directly apply RL to real-world environments. Due to the reality gap simulated environment does not match perfectly to the real-world scenario and additional learning cannot be performed. Therefore, an efficient approach is required for RL to find an optimal control policy and get better learning efficacy. In this paper, we propose federated reinforcement learning based on multi agent environment which applying a new federation policy. The new federation policy allows multi agents to perform learning and share their learning experiences with each other e.g., gradient and model parameters to increase their learning level. The Actor-Critic PPO algorithm is used with four types of RL simulation environments, OpenAI Gym's CartPole, MoutainCar, Acrobot, and Pendulum. In addition, we did real experiments with multiple Rotary Inverted Pendulum (RIP) to evaluate and compare the learning efficiency of the proposed scheme with both environments.

show abstract

“…A cyber physical system was used to deploy imitation reinforcement learning for simplifying the deep understanding in nonlinearities of an inverted pendulum. The method was applied for rotary inverted pendulum in open flow network [14].…”

Section: Introductionmentioning

confidence: 99%

Fuzzy Swing Up Control and Optimal State Feedback Stabilization for Self-Erecting Inverted Pendulum

2020

View full text Add to dashboard Cite

This paper presents the realisation of self-erecting inverted pendulum controls via two switched control approaches, a rule based fuzzy control for swing up inverted pendulum rod to pose upright position from downright position and an optimal state feedback control for stabilization as pendulum on upright position close to its equilibrium vertical line. The aim of this study is to solve two important problems on self-erecting inverted pendulum; swing up and stability in its upright balance position. Simulation and experimental results showed that control methods enabled the inverted pendulum swinging up and reaching its stable attitude in upright position even though small impulse and pulse disturbances were given. INDEX TERMS Fuzzy swing up, optimal state feedback stabilization, self erecting inverted pendulum.

show abstract

Imitation Reinforcement Learning-Based Remote Rotary Inverted Pendulum Control in OpenFlow Network

Cited by 27 publications

References 16 publications

Artificial Bee Colony Optimization Algorithm Incorporated With Fuzzy Theory for Real-Time Machine Learning Control of Articulated Robotic Manipulators

Artificial Bee Colony Optimization Algorithm Incorporated With Fuzzy Theory for Real-Time Machine Learning Control of Articulated Robotic Manipulators

Federated Reinforcement Learning Acceleration Method for Precise Control of Multiple Devices

Fuzzy Swing Up Control and Optimal State Feedback Stabilization for Self-Erecting Inverted Pendulum

Contact Info

Product

Resources

About