Silicon Ingot Pulling Control for Oxygen Concentration Precision
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a demand for improving the quality of single crystal silicon ingots, particularly in controlling the oxygen concentration during the production process.
Innovation Solution
A method for controlling a pulling apparatus that involves obtaining actual results data, generating an estimation model to predict oxygen concentration, adjusting operation amounts, and determining the operation amounts for the next batch to achieve target concentration, using a control program and apparatus to enhance the quality of single crystal silicon ingots.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If conventional production methods are used, then production efficiency is maintained, but oxygen concentration control precision deteriorates
Solution Approach 1:
The control system performs preliminary actions by generating an estimation model before production based on actual results data from previous batches. This model predicts oxygen concentration outcomes, allowing operators to adjust operation amounts in advance for the next batch, thereby improving oxygen concentration control precision without requiring complex real-time intervention systems.
Solution Approach 2:
The system implements feedback by obtaining actual results data (measured oxygen concentration and operation amounts) from each production batch and using this data to generate or update an estimation model. This closed-loop feedback mechanism enables continuous improvement of oxygen concentration control while maintaining manageable system complexity through iterative learning.
2Productivity
If production batch size is increased, then productivity improves, but quality consistency deteriorates
Solution Approach 1:
The system applies parameter changes by adjusting operation amounts (such as pulling speed, heating power, or atmospheric conditions) based on predictions from the estimation model. By systematically modifying these parameters according to the model's recommendations, the system maintains quality consistency even when production batch size increases, as each batch benefits from optimized parameter settings derived from accumulated data.
3Measurement precision
If more measurement data is collected, then measurement precision improves, but data processing complexity increases
Solution Approach 1:
The system uses copying by creating an estimation model that replicates the relationship between operation amounts and oxygen concentration outcomes based on historical data. Instead of directly processing all raw measurement data during production, the system copies the essential patterns into a predictive model, thereby maintaining measurement precision while reducing data processing complexity during actual production operations.
Data Source
AI summary
A method for controlling a pulling apparatus for a single crystal silicon ingot includes: obtaining actual results data that relates a measured value of oxygen concentration of a single crystal silicon ingot produced by the pulling apparatus and an operation amount of the pulling apparatus during production, generating an estimation model that estimates the oxygen concentration of a single crystal silicon ingot to be produced by the pulling apparatus based on the actual results data, adjusting the operation amount to be input to the estimation model so that an estimated value of the oxygen concentration of the single crystal silicon ingot by the estimation model becomes target concentration, and determining the adjusted operation amount as the operation amount for producing a single crystal silicon ingot in the next batch of the pulling apparatus.