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

2012 ◽  
Vol 2012 ◽  
pp. 1-9 ◽  
Author(s):  
Ying-Chung Wang ◽  
Chiang-Ju Chien
Keyword(s):  
Neural Network ◽  
Recurrent Neural Network ◽  
Iterative Learning Control ◽  
Design Method ◽  
Tracking Error ◽  
Learning Control ◽  
Iterative Learning ◽  
Learning Performance ◽  
Adaptive Laws ◽  
Control Objective

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.

A Two-Step Optimization-Based Iterative Learning Control for Quadrotor Unmanned Aerial Vehicles

2021 ◽  
Vol 143 (7) ◽  
Author(s):  
Revant Adlakha ◽  
Minghui Zheng
Keyword(s):  
Unmanned Aerial Vehicles ◽  
Design Problem ◽  
Iterative Learning Control ◽  
Design Method ◽  
Learning Control ◽  
Iterative Learning ◽  
Learning Performance ◽  
Aerial Vehicles ◽  
Quadrotor Unmanned Aerial Vehicles ◽  
Optimization Based Design

Abstract This paper presents a two-step optimization-based design method for iterative learning control and applies it onto the quadrotor unmanned aerial vehicles (UAVs) trajectory tracking problem. Iterative learning control aims to improve the tracking performance through learning from errors over iterations in repetitively operated systems. The tracking errors from previous iterations are injected into a learning filter and a robust filter to generate the learning signal. The design of the two filters usually involves nontrivial tuning work. This paper presents a new two-optimization design method for the iterative learning control, which is easy to obtain and implement. In particular, the learning filter design problem is transferred into a feedback controller design problem for a purposely constructed system, which is solved based on H-infinity optimal control theory thereafter. The robust filter is then obtained by solving an additional optimization to guarantee the learning convergence. Through the proposed design method, the learning performance is optimized and the system's stability is guaranteed. The proposed two-step optimization-based design method and the regarding iterative learning control algorithm are validated by both numerical and experimental studies.

Iterative Learning Control for Hybrid Systems

2020 ◽  
Author(s):  
Kirti D. Mishra ◽  
K. Srinivasan
Keyword(s):  
Hybrid Systems ◽  
Iterative Learning Control ◽  
Design Method ◽  
Tracking Error ◽  
Control Design ◽  
Learning Control ◽  
Iterative Learning ◽  
Linear Matrix ◽  
Triangular Structure ◽  
Form Representation

Abstract Iterative learning control (ILC) has been growing in applicability, along with growth in theory for classes of linear and nonlinear systems, and this study extends the theory of ILC to hybrid systems. A lifted form representation of hybrid systems with input-output dependent switching rules is developed, and the proposed lifted form representation is modeled as a switched system with arbitrary/unconstrained switching rules in the trial domain for control design. The causality of hybrid systems in the time domain results in the (lower) triangular structure of switched systems in the trial domain, the triangular structure enabling systematic and efficient control design. A unique aspect of the control design method developed for ILC of hybrid systems in this study is that a solution to the required set of linear matrix inequalities (LMIs) is guaranteed to exist under mild assumptions, which is in contrast to many other studies proposing LMI based solutions in controls literature. The proposed method is validated numerically for a motion control application, and robust and monotonic convergence of the tracking error to zero is demonstrated.

Backstepping Adaptive Iterative Learning Control for Robotic Systems

2013 ◽  
Vol 284-287 ◽  
pp. 1759-1763
Author(s):  
Ying Chung Wang ◽  
Chiang Ju Chien ◽  
Chi Nan Chuang
Keyword(s):  
Neural Network ◽  
Iterative Learning Control ◽  
Fuzzy Neural Network ◽  
Tracking Error ◽  
Learning Control ◽  
Iterative Learning ◽  
Robotic Systems ◽  
Adaptive Iterative Learning Control ◽  
Repetitive Tasks ◽  
Fuzzy Neural

A backstepping adaptive iterative learning control for robotic systems with repetitive tasks is proposed in this paper. The backstepping-like procedure is introduced to design the AILC. A fuzzy neural network is applied for compensation of the unknown certainty equivalent controller. Using a Lyapunov like analysis, we show that the adjustable parameters and internal signals remain bounded, the tracking error will asymptotically converge to zero as iteration goes to infinity.

An Output Based Adaptive Iterative Learning Control with Particle Swarm Optimization for Robotic Systems

2013 ◽  
Vol 479-480 ◽  
pp. 737-741
Author(s):  
Ying Chung Wang ◽  
Chiang Ju Chien ◽  
Chi Nan Chuang
Keyword(s):  
Particle Swarm Optimization ◽  
Iterative Learning Control ◽  
Tracking Error ◽  
Particle Swarm ◽  
Learning Control ◽  
Iterative Learning ◽  
Learning Performance ◽  
Robotic Systems ◽  
Adaptive Iterative Learning Control ◽  
Swarm Optimization

We consider an output based adaptive iterative learning control (AILC) for robotic systems with repetitive tasks in this paper. Since the joint velocities are not measurable, a sliding window of measurements and an averaging filter approach are used to design the AILC. Besides, the particle swarm optimization (PSO) is used to adjust the learning gains in the learning process to improve the learning performance. Finally, a Lyapunov like analysis is applied to show that the norm of output tracking error will asymptotically converge to a tunable residual set as iteration goes to infinity.

Iterative learning control algorithm for sensor fault nonlinear systems

2020 ◽  
pp. 1-8
Author(s):  
Zimian Lan
Keyword(s):  
Nonlinear Systems ◽  
Iterative Learning Control ◽  
Control Algorithm ◽  
Tracking Error ◽  
Learning Control ◽  
Open Loop ◽  
Iterative Learning ◽  
Sensor Fault ◽  
Initial State ◽  
Control Gain

In this paper, we propose a new iterative learning control algorithm for sensor faults in nonlinear systems. The algorithm does not depend on the initial value of the system and is combined with the open-loop D-type iterative learning law. We design a period that shortens as the number of iterations increases. During this period, the controller corrects the state deviation, so that the system tracking error converges to the boundary unrelated to the initial state error, which is determined only by the system’s uncertainty and interference. Furthermore, based on the λ norm theory, the appropriate control gain is selected to suppress the tracking error caused by the sensor fault, and the uniform convergence of the control algorithm and the boundedness of the error are proved. The simulation results of the speed control of the injection molding machine system verify the effectiveness of the algorithm.

A data-driven approach for trajectory-based aircraft operation with controlled time of arrival and along-track wind effects

2020 ◽  
Vol 42 (12) ◽  
pp. 2166-2177
Author(s):  
Gaoyang Jiang ◽  
Zhongsheng Hou
Keyword(s):  
Control Problem ◽  
Iterative Learning Control ◽  
Tracking Error ◽  
Learning Control ◽  
Iterative Learning ◽  
Data Driven ◽  
Time Of Arrival ◽  
Multiple Time ◽  
Data Driven Approach ◽  
Along Track

Trajectory-based aircraft operation and control is one of the hot issues in air traffic management. However, the accurate mechanism modeling of aircraft is tough work, and the operation data have not been effectively utilized in many studies. So, in this work, we apply the model-free adaptive iterative learning control method to address the time-of-arrival control problem in trajectory-based aircraft operation. This problem is first formulated into a trajectory tracking problem with along-track wind disturbance. Through rigorous analysis, it is shown that this method, combined with point-to-point iterative learning control (ILC) strategy, can effectively deal with the arrival time control problem with multiple time constraints. Then, the terminal ILC strategy is applied, aiming to resolve the same problem with a time constraint at the end point. Compared with the PID (Proportional Integral Derivative) type ILC, the proposed method improves control performance by 11.15% in root mean square of tracking error and 9.32% in integral time absolute error. The sensitivity and flexibility of the data-driven approach is further verified through numerical simulations.

Sign in / Sign up

Export Citation Format

Share Document