2023•Unpublished venueRequires access

Model Free iterative learning for table motion control of lithography machine

Yuting Li, Ping Luo, Yunfeng Peng, Zhiyuan Liu

Open publisher page 2 citations

Abstract

In this paper, for photomask-free lithography machines, table motion control, as an important module of lithography equipment, is equipped with wafers to complete the exposure motion. It can be seen that the table motion trajectory has the characteristics of repeatability and periodicity when it is exposed in each exposure field. The application of iterative learning control has a good control effect on repetitive motion. However, due to the uncertainties and time-varying problems of the table system model of lithography machine, a model free iterative learning method is proposed also called data-driven iterative learning strategy. The learning law is calculated by using the input and output data of the system last time and does not depend on the system model. Model Free iterative learning control is applied to the workpiece control system, and model Free iterative learning control is regarded as feedforward control. Model Free iterative learning algorithm can calculate the input signal of the system according to the expected trajectory of the system, the last output and input of the system, and the learning law of the iterative learning control adopts offline learning mode. Finally, simulation verification is carried out, and the comparison between model-free iterative learning and other traditional control algorithms shows that the tracking error decreases with the increase of iteration. Compared with the traditional iterative learning control algorithm, it has better robustness and improves the precision of the control system.

About this research paper

What this paper is about

In this paper, for photomask-free lithography machines, table motion control, as an important module of lithography equipment, is equipped with wafers to complete the exposure motion. It can be seen that the table motion trajectory has the characteristics of repeatability and periodicity when it is exposed in each exposure field. The application of iterative learning control has a good control effect on repetitive motion. However, due to the uncertainties and time-varying problems of the table system model of lithography machine, a model free iterative learning method is proposed also called data-driven iterative learning strategy. The learning law is calculated by using the input and output data of the system last time and does not depend on the system model. Model Free iterative learning control is applied to the workpiece control system, and model Free iterative learning control is regarded as feedforward control. Model Free iterative learning algorithm can calculate the input signal of the system according to the expected trajectory of the system, the last output and input of the system, and the learning law of the iterative learning control adopts offline learning mode. Finally, simulation verification is carried out, and the comparison between model-free iterative learning and other traditional control algorithms shows that the tracking error decreases with the increase of iteration. Compared with the traditional iterative learning control algorithm, it has better robustness and improves the precision of the control system.

Why it matters

OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

In this paper, for photomask-free lithography machines, table motion control, as an important module of lithography equipment, is equipped with wafers to complete the exposure motion. It can be seen that the table motion trajectory has the characteristics of repeatability and periodicity when it is exposed in each exposure field. The application of iterative learning control has a good control effect on repetitive motion. However, due to the uncertainties and time-varying problems of the table system model of lithography machine, a model free iterative learning method is proposed also called data-driven iterative learning strategy. The learning law is calculated by using the input and output data of the system last time and does not depend on the system model. Model Free iterative learning control is applied to the workpiece control system, and model Free iterative learning control is regarded as feedforward control. Model Free iterative learning algorithm can calculate the input signal of the system according to the expected trajectory of the system, the last output and input of the system, and the learning law of the iterative learning control adopts offline learning mode. Finally, simulation verification is carried out, and the comparison between model-free iterative learning and other traditional control algorithms shows that the tracking error decreases with the increase of iteration. Compared with the traditional iterative learning control algorithm, it has better robustness and improves the precision of the control system.

Key concepts: Iterative learning control, Computer science, Iterative method, Motion control, Robustness (evolution), Trajectory, Feed forward, Motion system

Related papers

Back to paper searchBrowse research topicsOriginal source
Model Free iterative learning for table motion control of lithography machine — Research Paper | ScholarLens