Abstract:Run-to-Run (R2R) control has been widely used in semiconductor manufacturing to compensate for process disturbance and to improve quality. The traditional R2R controller only takes the process output in a previous run as its input and generates an optimal recipe for the next run. In a multistage semiconductor manufacturing process, variations in upstream stations are propagated to downstream stations. However, the information from upstream stations is not considered by existing controllers. In addition, most R2R processes have a limited capacity; the products must be processed in batches. Therefore, if the incoming materials could be grouped with small within-batch variations and large batch-to-batch variations and the recipes are customized for each batch to drive all batch averages toward the same target value, the output variation could be reduced and quality improved. A batch Exponentially Weighted Moving Average (EWMA) controller is proposed in this article. It employs a modified K-means algorithm to group the incoming materials into batches with a fixed and equal size while minimizing the within-batch variation. The controller then generates the control settings by taking both the batch information and the feedback quality information into account. Simulation studies show that the proposed controller could significantly reduce output variation and improve quality.

Adaptive EWMA controller based on dynamic modeling in semiconductor manufacturing

Model Quality Evaluation in Semiconductor Manufacturing Process with Ewma Run-To-Run Control

Performance Analysis of Single EWMA Controller Subject to Metrology Delay under Dynamic Models

An Engineering-Effective Adaptive Model Algorithm Control (AMAC) and Simulation

Self-Evolving Offset-Free Model Predictive Control with Model-Plant Mismatch for Dynamic Working-Point Change Tasks in Industrial Processes

Stability analysis of semiconductor manufacturing process with EWMA run-to-run controllers

Performance Analysis of Double EWMA Controller under Dynamic Models with Drift

On Variable Weight Ewma Based Run-to-Run Control in Mixed Product Process

Performance analysis and comparison between two forms of double EWMA controllers in industrial process

Learning MPC for Process Dynamic Working Condition Change Tasks under Model Mismatch

EWMA for Run-to-run Control with Time-Varying Discount Factor in High-Mix Process

Adaptive Sliding Mode Control of the A-axis Used for Blisk Manufacturing

MIMO EWMA-CUSUM Condition-Based Statistical Process Control in Manufacturing Processes.

Fuzzy predictive R2R control to CMP process

An Analysis and Performance Assessment of Double Ewma Control Systems

Cycle Forecasting EWMA (CF-EWMA) Approach for Drift and Fault in Mixed-Product Run-to-run Process

Grey model based immune predictive R2R control of CMP process

Comments on “extensions and Performance/Robustness Tradeoffs of the EWMA Run-to-Run Controller by Using the Internal Model Control Structure”

Online Sequential Extreme Learning Machine Based Adaptive Control for Wastewater Treatment Plant

A machine learning based EMA-DCPM algorithm for production scheduling

A Batch-Based Run-To-Run Process Control Scheme for Semiconductor Manufacturing