PCA Model for Substrate Processing Fault Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Monitoring and fault detection in manufacturing processes are time-consuming due to the large number of varying data parameters, necessitating a method to reduce the load and enhance sensitivity and control.

Innovation Solution

Grouping processing parameters into sub-groups of correlated parameters and applying different weighting factors when constructing a PCA model to improve sensitivity and control in fault detection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional monitoring methods are used to track all processing parameters, then comprehensive fault detection coverage is achieved, but monitoring time and computational load increase significantly

Engineering Contradiction:

Improvefault detection coverageVSAvoidmonitoring time

Core Design Contradiction:

ReliabilityVSLoss of time

Solution Approach 1:

The patent segments the monitoring system by grouping correlated parameters into sub-groups and constructing separate PCA models for each sub-group. This segmentation reduces the computational complexity compared to monitoring all parameters uniformly, while maintaining comprehensive fault detection coverage through the collective analysis of multiple sub-groups.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces PCA models as intermediary tools that transform raw processing parameters into statistical quantities (T² and Q statistics). These statistical quantities serve as mediators that condense information from multiple parameters into fewer indicators, reducing monitoring time while preserving fault detection capability.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Device complexity

If uniform weighting is applied to all parameters in PCA model construction, then simplicity is maintained, but sensitivity to specific faults is reduced

Engineering Contradiction:

Improvemodel construction simplicityVSAvoidfault detection sensitivity

Core Design Contradiction:

Device complexityVSMeasurement precision

Solution Approach 1:

The patent applies local quality by assigning different weighting factors to different parameters within the same sub-group based on their individual importance. This allows the model to give more attention to critical parameters that are more indicative of specific faults, thereby enhancing fault detection sensitivity without significantly increasing overall model complexity.

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

The patent changes the parameter weighting scheme from uniform to differentiated weighting factors. By adjusting the weighting factors according to parameter importance, the PCA model achieves higher sensitivity to specific faults while maintaining a relatively simple construction process through systematic weighting assignments.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS10606253B2Method of monitoring processing system for processing substrate

Publication Date: 2020.03.31 UNITED MICROELECTRONICS CORP

AI summary

A method of monitoring a processing system for processing a substrate is provided. The method includes the following steps: acquiring data from the processing system for a plurality of parameters, the data including a plurality of data values; grouping the parameters into a plurality of sub-groups, each of the sub-groups including a plurality of correlated parameters; constructing a principle components analysis (PCA) model from the data values for the correlated parameters in a first one of the sub-groups, including normalizing the data values in the first one of the sub-groups with a first weighting factor and a second weighting factor, wherein the first weighting factor is different from the second weighting factor; and determining a statistical quantity using the PCA model.