Adaptive Dynamic Programming-based Decentralized Sliding Mode Optimal Control for Modular and Reconfigurable Robots

An, Tianjiao; Men, Xiaodong; Xinyao, Zhang; Li, Yuanchun; Dong, Bo

doi:10.1109/ssci44817.2019.9002953

Cited by 3 publications

(3 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Unlike the convention modeling method (An et al, 2019; Li et al, 2020), we have combined the unmodeled dynamic terms and IDC terms together, which deem as a disturbance torque. Then, according to the zero-sum game theory, the optimal control policy and worst disturbance policy are obtained by the HJI function, respectively, that no need extra sub-controller which can reduce the computational burden.…”

Section: Reconfigurable Robots’ Dynamic Modelmentioning

confidence: 99%

“…Optimal control method has been proposed nearly 70 years ago (Bellman and Dreyfus, 2013; Pontryagin, 1959), and plenty of literatures are recorded (An et al, 2021; An et al, 2021; Wei et al, 2021). Adaptive dynamic programming (ADP) theory is a key direction to undertake approximate optimal control system issues of discrete-time (Hu et al, 2021; Luo et al, 2020; Zhu et al, 2020a), continuous-time (An et al, 2021; Shan et al, 2020; Wei et al, 2020), data driven (Gao et al, 2018; Li et al, 2020; Su et al, 2020), and furthermore robot systems with input/output constraints (An et al, 2020; Lu et al, 2020; Ren et al, 2019), external disturbance (An et al, 2019; Song and Lewis, 2020; Xia et al, 2020), actuator failures (Dai et al, 2018; Jiao et al, 2018; Ma et al, 2020), and so on. Kong et al (2021b) developed n -link manipulator approximate optimal law via saturation nonlinearity.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Decentralized position–force zero-sum approximate optimal control for reconfigurable robots with unmodeled dynamic

Zhu

Wang

2022

Transactions of the Institute of Measurement and Control

Self Cite

View full text Add to dashboard Cite

In this paper, the position–force–based approximate optimal control method is developed for reconfigurable robots using zero-sum game strategy. By utilizing the Newton–Euler iteration technique, the robotic system’s dynamic model is formulated and the state space equation is derived. According to adaptive dynamic programming (ADP) and neural network algorithm, the trajectory tracking control problem is transformed into a zero-sum game-based optimal control issue. The optimal control policy and worst disturbance policy are obtained by Hamilton–Jacobi–Issacs (HJI) function, respectively. Unlike the conventional learning–based robotic control method, the proposed zero-sum game-based method no need extra sub-controller that can reduce the computational burden. The reconfigurable robot system’s tracking error is uniformly ultimately bounded by the Lyapunov theorem. Finally, simulation experiments demonstrate the advantages of the proposed method.

show abstract

Section: Reconfigurable Robots’ Dynamic Modelmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Decentralized position–force zero-sum approximate optimal control for reconfigurable robots with unmodeled dynamic

Zhu

Wang

2022

Transactions of the Institute of Measurement and Control

Self Cite

View full text Add to dashboard Cite

show abstract

“…In the first case, structure control is embedded in a single device [6]. In the second case, the controller can be any module in the structure [7,8]. In the third case, the controller incorporates features of the previous controllers; that is, there is a central controller that sends parameters to the modules that translate said information to perform a task [9][10][11].…”

Section: Introductionmentioning

confidence: 99%

Artificial Neural Network Controller for a Modular Robot Using a Software Defined Radio Communication System

2020

View full text Add to dashboard Cite

Modular robots are flexible structures that offer versatility and configuration options for carrying out different types of movements; however, disconnection problems between the modules can lead to the loss of information, and, therefore, the proposed displacement objectives are not met. This work proposes the control of a chain-type modular robot using an artificial neural network (ANN) that enables the robot to go through different environments. The main contribution of this research is that it uses a software defined radio (SDR) system, where the Wi-Fi channel with the best signal-to-noise Ratio (SNR) is selected to send the information regarding the simulated movement parameters and obtained by the controller to the modular robot. This allows for faster communication with fewer errors. In case of a disconnection, these parameters are stored in the simulator, so they can be sent again, which increases the tolerance to communication failures. Additionally, the robot sends information about the average angular velocity, which is stored in the cloud. The errors in the ANN controller results, in terms of the traveled distance and time estimated by the simulator, are less than 6% of the real robot values.

show abstract

Design and Implementation of Fractional Order Controllers Using Nelder-Mead Algorithm

Naithani

Chaturvedi

Juneja³

et al. 2021

2021 International Conference on Computational Performance Evaluation (ComPE)

View full text Add to dashboard Cite

Adaptive Dynamic Programming-based Decentralized Sliding Mode Optimal Control for Modular and Reconfigurable Robots

Cited by 3 publications

References 13 publications

Decentralized position–force zero-sum approximate optimal control for reconfigurable robots with unmodeled dynamic

Decentralized position–force zero-sum approximate optimal control for reconfigurable robots with unmodeled dynamic

Artificial Neural Network Controller for a Modular Robot Using a Software Defined Radio Communication System

Design and Implementation of Fractional Order Controllers Using Nelder-Mead Algorithm

Contact Info

Product

Resources

About