2019.05.28
Overview
An optimization method based on feature extraction and machine learning is provided. At least one input parameter is received. A plurality of first historical mold data is retrieved. Similarity calculation is performed based on the input parameter and the first historical mold data. Multiple candidate mold data is selected according to the similarity calculation. The mold design parameters corresponding to each input parameter in the candidate mold data are replaced by each input parameter, and a plurality of first representative mold materials for performing the first simulation analysis are generated. A variety of key feature parameters is found, and multiple second historical mold data is retrieved based on these key feature parameters. An expected data is found and the mold design parameters in the expected data is filtered and optimized to find second representative mold materials for performing the second simulation analysis. At least one mold production parameter is generated.Category
發明
Patented
16/424,168
US11,182,519B2
Filing Date
2019.05.28Expired Date
2040.04.16Notification
2020.10.20