An Initial Residual Stress Inference Method by Incorporating Monitoring Data and Mechanism Model

Abstract Initial residual stress is the main reason causing machining deformation of workpiece, which has been deemed as one most important aspects quality issues. The inference distribution initial inside blank significant meaning for control. Due to principle error existing detection methods, there are still challenges in practical applications. Aiming at problem field, an method by incorporating monitoring data and mechanism model proposed this paper. Monitoring during process used represent macroscopic characterization unbalanced stress, finite element numerical so solve that analytic difficult establish; policy gradient approach introduced descent combination learning model. Finally, field obtained through iterative calculation based on fusing Verification results show can accurately effectively reflect actual process.