A Vibration Control Method for Hybrid-Structured Flexible Manipulator Based on Sliding Mode Control and Reinforcement Learning

Long, Teng; Li, En; Hu, Yunqing; Yang, Lei; Fan, Junfeng; Liang, Zize; Guo, Rui

doi:10.1109/tnnls.2020.2979600

Cited by 45 publications

(20 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ye et al (2022a) presented a saturated finite-time fast terminal sliding mode controller, while the Legendre polynomial-based neural network was incorporated to approximate the unknown nonlinear dynamics. Long et al (2021) used the improved SMC based on the nominal model as the main controller, and adopted the reinforcement learning controller based on the actor-critic to output the compensation torque to suppress the end vibration of a flexible manipulator with hybrid structure. It can be noted that the aforementioned neural network is used as a compensator to approximate uncertainties or enhance control signals.…”

Section: Introductionmentioning

confidence: 99%

Vibration control of a translational coupled double flexible beam system using sliding mode neural network fuzzy control

Qiu

Chen

2022

Transactions of the Institute of Measurement and Control

View full text Add to dashboard Cite

A sliding mode neural network fuzzy control (SMNNFC) method is investigated to suppress the vibration of a translational coupled double flexible beam system, equipped with an AC servomotor and several piezoelectric actuators. Adjacent beams and slider moving frame are connected at the tip by elastic springs. Based on the finite element method, the system model is established to recognize the vibration characteristics. Furthermore, two laser displacement sensors are used to decouple the first two bending modes of the double flexible beam system. In the applied SMNNFC strategy, a neural-fuzzy framework is designed to obtain robust control performance and alleviate the chattering phenomenon. Considering unknown and varying system uncertainty, a parameter updating algorithm is adopted. The stability of SMNNFC is analyzed. The experimental setup is constructed and experiments are conducted, including set-point vibration control and simultaneous translation and vibration control under trapezoidal and sinusoidal trajectories. The experimental results demonstrate that the SMNNFC scheme has advantages in suppressing both the large and low amplitude vibrations of the coupling double flexible beam system.

show abstract

Section: Introductionmentioning

confidence: 99%

Vibration control of a translational coupled double flexible beam system using sliding mode neural network fuzzy control

Qiu

Chen

2022

Transactions of the Institute of Measurement and Control

View full text Add to dashboard Cite

show abstract

“…Pinto et al (2017) realized several tasks by using actor-critic algorithm and asymmetric input, such as picking, pushing, and moving blocks of real robots. Long et al (2021) used a RL controller based on actor-critic structure as an auxiliary controller to compensate for the output torque, and verified its effectiveness and robustness. He et al (2021) used RL control based on actor-critic structure to suppress the vibration of a flexible two-link manipulator while retaining trajectory tracking.…”

Section: Introductionmentioning

confidence: 99%

Vibration control of three coupled flexible beams using reinforcement learning algorithm based on proximal policy optimization

Qiu

Zhang

2022

Journal of Intelligent Material Systems and Structures

View full text Add to dashboard Cite

In order to suppress the vibration of a class of multi-coupling flexible beams, a three-coupling flexible beam experimental device is constructed. The vibration characteristics and active vibration control algorithms of three coupled flexible beams are studied. Due to the complex residual vibration of multi-coupled flexible beam structure, which shows the interaction between beams and close frequency characteristics, the dynamic model of coupled flexible beam structure is established by finite element method (FEM) and its theoretical modal analysis is carried out. Furthermore, the parameters of the experimental system are identified, and the close modal frequencies and damping ratios of the actual system are obtained. The characteristics of close modal vibration are verified. A reinforcement learning (RL) vibration controller based on proximal policy optimization (PPO) algorithm is designed. The algorithm mainly includes actor neural network and critic neural network, which are used to approximate the control strategy and evaluate state respectively. Based on theoretical modeling and experimental identification correction model, the agent learns offline, and then transfers the learned policy parameters to the experimental system for vibration control experiments. The experimental results demonstrate that the designed RL controller can effectively suppress the residual vibration of three coupled flexible beams.

show abstract

“…Te essential features of neural networks are their way to generate good models of nonlinear systems, their decentralized and multithreaded structure, which renders neural-based control methods quicker, their ease of implementation by software or hardware, and their capacity to learn and adapt to the practices of any real process. In the infuence of environmental disturbances or when the IFO drive system experiences defective decoupling due to rotor time constant variations, fuzzy controllers have been shown to enhance measurement accuracy [10][11][12]. Because fuzzy logic and neural networks have better tracking properties than traditional controllers, they are getting popular as estimators and controllers for a wide range of industrial applications.…”

Section: Introductionmentioning

confidence: 99%

Analysis and Implementation of Sliding Mode Controller-Based Variable Frequency Drive Using the SCADA System

Aljafari

Kumar

Indragandhi

et al. 2022

International Transactions on Electrical Energy Systems

View full text Add to dashboard Cite

Vector control of an asynchronous machine is traditionally accomplished by analogizing it to a separately excited DC machine. It provides decoupled torque and flux control that is perpendicular to each other, ensuring that neither vector interferes with the other. So, despite their close interconnection, torque and speed control are accomplished separately. The rotor flux is aligned with the direct axis of the synchronously rotating reference frame to achieve this. The PI controllers are critical in achieving the variable frequency drive (VFD) desired topology. The system employs three types of controllers: flux, speed, and torque. The flux controller is easy to tune, but the speed and torque controllers are more difficult to tune because the speed controller's output is the torque controller’s reference signal. Furthermore, there is no well-defined method for tuning the controllers in a vector control system. However, perfect tuning is required for the machine’s better dynamic behavior. It is clear from the above analysis that system identification is critical for tuning PI controllers. However, as an asynchronous machine, obtaining a decoupled system transfer function is extremely difficult. To solve this problem, the proposed system combines a seven-level pulse width modulation (PWM) inverter for vector control of a three-phase asynchronous nonstandard induction machine used in critical applications in nuclear power plants with a sliding mode control technique that eliminates the complexity of PI tuning. A second-order sliding mode controller could be used in the future to reduce the chattering and parameter variation effects. This controller can be enhanced with fuzzy logic principles to make it more robust and reliable, allowing it to be used in future drive designs for high-rating motors with critical applications.

show abstract

A Vibration Control Method for Hybrid-Structured Flexible Manipulator Based on Sliding Mode Control and Reinforcement Learning

Cited by 45 publications

References 28 publications

Vibration control of a translational coupled double flexible beam system using sliding mode neural network fuzzy control

Vibration control of a translational coupled double flexible beam system using sliding mode neural network fuzzy control

Vibration control of three coupled flexible beams using reinforcement learning algorithm based on proximal policy optimization

Analysis and Implementation of Sliding Mode Controller-Based Variable Frequency Drive Using the SCADA System

Contact Info

Product

Resources

About