Evolutionary Algorithm for Plasma Etch Process Change Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting process changes in plasma etch systems, such as in the semiconductor industry, face limitations due to reliance on manual selection of data channels and scaling factors, which are time-consuming and often result in suboptimal signal-to-noise ratios, especially when using multiple data channels.
Innovation Solution
The implementation of an evolutionary computing technique, specifically a genetic algorithm, to generate optimized process state functions and scaling coefficients for detecting process changes by analyzing datasets from plasma etch processes, allowing for automatic selection and optimization of data channels and scaling factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual selection of data channels and scaling factors is used, then user control and interpretability are maintained, but the signal-to-noise ratio for process change detection becomes suboptimal and the process becomes time-consuming
Solution Approach 1:
The system performs automatic optimization of the process state function without requiring manual user intervention. The evolutionary algorithm autonomously selects data channels and determines scaling coefficients by evaluating process data and detecting changes, thereby eliminating the time-consuming manual selection process while achieving optimal signal-to-noise ratio
Solution Approach 2:
The manual mechanical process of selecting and tuning data channels and scaling factors is replaced by an automated computational system using evolutionary algorithms. This substitution transforms the manual optimization process into an automated computational search that efficiently identifies optimal parameters without human intervention
2Measurement precision
If multiple data channels are used to increase detection sensitivity, then the signal-to-noise ratio improves, but the complexity of the process state function increases
Solution Approach 1:
The evolutionary algorithm extracts and identifies only the most relevant data channels from the available spectrum, eliminating redundant or noisy channels. This extraction process reduces the number of channels that need to be monitored while maintaining or improving detection sensitivity, thereby simplifying the process state function
Solution Approach 2:
The algorithm segments the available data channels by evaluating their individual contribution to process change detection. Through this segmentation process, channels are grouped or selected based on their relevance, allowing the system to focus on a optimized subset of channels that provides maximum detection sensitivity with reduced complexity
3Measurement precision
If evolutionary computing techniques are used to automatically optimize process state functions, then detection sensitivity and signal-to-noise ratio improve, but computational complexity and processing time increase
Solution Approach 1:
The evolutionary algorithm performs partial optimization by focusing computational effort on the most critical parameters and data channels rather than exhaustively optimizing all possible parameters. This partial action approach achieves sufficient detection sensitivity without requiring complete optimization of every aspect of the process state function, thereby reducing computational complexity
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The present invention provides a method for creating a process change detection algorithm. An evolutionary computing technique is applied to at least one process dataset containing at least one known process change. The evolutionary computing technique will generate a process state function (or a scaling coefficient set for use with an existing process state function) that optimizes detection of the known process changes. The generated process state function or coefficients can then be applied thereafter to future datasets (either in real-time or after processing) to detect process changes.