Sampling-Function Clustering for Faster Parameter Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing parameter optimization methods are time-consuming and inefficient, as they require multiple iterations to achieve a process target value.
Innovation Solution
A parameter optimization device and method that includes data acquisition, sampling function calculation, clustering, and parameter recommendation modules to efficiently group and recommend parameter values, reducing the number of measurement batches and iterations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional parameter optimization method is used (one process parameter output at a time with sequential measurement), then the method is simple to implement, but the time consumption and iteration count increase significantly
Solution Approach 1:
The patent segments the parameter optimization process into distinct functional modules: data acquisition module for collecting input-output pairs, sampling function calculation module for generating candidate parameters, clustering module for grouping similar parameters, and recommendation module for selecting optimal parameters. This segmentation enables parallel processing of multiple parameters simultaneously, dramatically reducing iteration time and improving optimization efficiency compared to sequential methods.
2Manufacturing precision
If multiple iterations are performed to achieve process target value, then the optimization accuracy improves, but the number of measurement batches increases
Solution Approach 1:
The patent performs preliminary actions by pre-calculating sampling function values for multiple candidate parameters before actual measurement. The sampling function calculation module generates probability distributions and identifies promising parameter ranges in advance, allowing the system to prioritize which parameters to measure first. This preliminary analysis reduces the number of measurement batches needed to achieve the process target value, as measurements are strategically directed toward most promising candidates rather than random or sequential exploration.
Solution Approach 2:
The patent implements feedback mechanisms where measurement results from previous iterations are fed back into the clustering and recommendation modules. The system continuously updates parameter groups based on observed outcomes, refining its understanding of the input-output relationship. This feedback loop enables the recommendation module to progressively identify optimal parameters with fewer iterations, improving both manufacturing precision and measurement efficiency by learning from each measurement batch.
Data Source
AI summary
A parameter optimization device includes a data acquisition module, a sampling function calculation module, a clustering module and a parameter recommendation module. The data acquisition module is configured to acquire several input parameter values and corresponding several measurement output values. The sampling function calculation module is configured to obtain several sampling function values according to the input parameter values and the measurement output values. The clustering module is configured to obtain several parameter groups according to the input parameter values and the sampling function values. The parameter recommendation module is configured to obtain several recommended parameter values from at least one of the parameter groups.