An Output-Recurrent-Neural-Network-Based Iterative Learning Control for Unknown Nonlinear Dynamic Plants

Abstract

We present a design method for iterative learning control system by using an output recurrent neural network (ORNN). Two ORNNs are employed to design the learning control structure. The first ORNN, which is called the output recurrent neural controller (ORNC), is used as an iterative learning controller to achieve the learning control objective. To guarantee the convergence of learning error, some information of plant sensitivity is required to design a suitable adaptive law for the ORNC. Hence, a second ORNN, which is called the output recurrent neural identifier (ORNI), is used as an identifier to provide the required information. All the weights of ORNC and ORNI will be tuned during the control iteration and identification process, respectively, in order to achieve a desired learning performance. The adaptive laws for the weights of ORNC and ORNI and the analysis of learning performances are determined via a Lyapunov like analysis. It is shown that the identification error will asymptotically converge to zero and repetitive output tracking error will asymptotically converge to zero except the initial resetting error.

Neural iterative learning identifier‐based iterative learning controller for time‐varying nonlinear systems

“…The neural network is trained to reach optimal weights at each iteration of the control algorithm, and updating of neural network weights is also iterative process. A neural controller is presented in several studies [15][16][17], and the training process of this neural controller needs a precise estimation of the system that is achieved using another neural network. In the approaches presented in several studies [14][15][16][17], the control iteration and identification process iteration are different; therefore, these approaches involve much computing that takes a lot of time.…”

“…A neural controller is presented in several studies [15][16][17], and the training process of this neural controller needs a precise estimation of the system that is achieved using another neural network. In the approaches presented in several studies [14][15][16][17], the control iteration and identification process iteration are different; therefore, these approaches involve much computing that takes a lot of time. As well, in the method presented in Patan et al and Patan and Patanl [15,16], the convergence is not monotonic; therefore, a law pass Q-filter must be used to achieve the monotonic convergence.…”

“…Unknown time-varying nonlinear systems [26] include a wide range of complex dynamical systems such as space weather system, neuroimaging, finance, the solar-terrestrial, and stem cell dynamics [27,28]. So far, few ILC approaches [15][16][17][19][20][21][22][23][24] have been presented particular for unknown discrete time-varying nonlinear systems. In all the research mentioned above, there are defects; therefore, in this paper, we propose a new approach in which these defects are rectified.…”

“…1. In this paper, the ILI and ILC algorithms are integrated (simultaneous) at each iteration; therefore, the proposed NILI-based ILC algorithm costs less time and effort than the algorithms presented in several studies [14][15][16][17]. 2.…”

Design of Iterative Learning Control Method with Global Convergence Property for Nonlinear Systems

“…The objective of iterative learning control is to overcome the imperfect knowledge of system structure to improve tracking performance as few trials as possible. Since ILC issue is originally proposed by Arimoto et al [1], applications of ILC can be widely found in industrial robot manipulator, chemical batch process, some medical equipment, manufacturing, and so forth [2][3][4][5][6][7].…”