A novel event-triggered adaptive tracking control framework for a manipulator with aperiodic neural network estimation
Article publication date: 7 June 2022
Issue publication date: 19 July 2022
The purpose of this study is developing the minimum parameter learning law for the weight updating, which reduces the updating of neural network (NN) weight only at triggering instants and makes a trade-off between the estimation accuracy and triggering frequency such that the computing complexity can be decreased. Besides that, a novel “soft” method is first constructed for the control updating at the triggered instants, to reduce the chattering effect of discontinued renewal of control. Addressing to the proposed control and updating method, a novel dead-zone condition with variable boundary about the triggered control signal is derived to ensure the positivity of adjacent execution intervals.
In this paper, to achieve the motion tracking of manipulator with uncertainty of system dynamics and the communication constraints in the control-execution channel, an adaptive event-triggered controller with NN identification is constructed to improve the transmission efficiency of control on the premise of the guaranteed performance. In the proposed method, the NN with intermittent updating is proposed to perform the uncertain approximation with the saved computation, and the triggered mechanism is constructed to regulate the transportation of the signal in the channel of controller-to-actuator.
According to the impulsive Lyapunov function, it can be proved that all the signals are semi-global uniformly ultimately bounded, and the positivity of adjacent execution intervals is also guaranteed by the proposed method. In addition, the chattering effect of control updating at the jumping instants can be relieved by the proposed “soft” mechanism, such that the control accuracy and stability can be guaranteed. Experiments on the JACO2 real manipulator are carried out to verify the effectiveness of the proposed scheme.
To the best of the author’s knowledge, this study is firstly to propose a “soft” method to reduce the chattering effect caused by discontinuous updating. Addressing to the updating method designed above, a novel dead-zone condition with variable threshold and boundary is first constructed to ensure the positivity of execution intervals.
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Gao, J. (2022), "A novel event-triggered adaptive tracking control framework for a manipulator with aperiodic neural network estimation", Assembly Automation, Vol. 42 No. 4, pp. 411-426. https://doi.org/10.1108/AA-02-2022-0025
Emerald Publishing Limited
Copyright © 2022, Emerald Publishing Limited