Iterative learning-based deep hole working depth optimization method

Abstract

The invention discloses an iterative learning-based deep hole working depth optimization method, which is used for solving the technical problem of poor practicality of an existing deep hole working depth optimization method. According to the technical scheme, the method comprises the steps of firstly building a cutting force prediction model based on a classic cutting force empirical formula; secondly in a deep hole working circulation process, correcting a model coefficient according to actually measured drilling force data when a working circulation is finished each time, and updating a predicted working depth, thereby realizing an iterative learning process of the model, and continuously improving the prediction precision of the model; and finally optimizing the working depth by utilizing the model after the learning to obtain a final optimization result. The model coefficient is corrected in real time through an iterative learning method by utilizing the deep hole working circulation, so that the prediction precision of the model is continuously improved, the working process is kept consistent with a real working process, and the problem of prediction precision insufficiency of the existing analysis model-based deep hole working depth optimization method is solved; and the practicality is high.

Description

technical field [0001] The invention relates to a deep hole machining depth optimization method, in particular to a deep hole machining depth optimization method based on iterative learning. Background technique [0002] The document "Modeling Chip-Evacuation Forces and Prediction of Chip-Clogging in Drilling, Journal of Manufacturing Science & Engineering, 2002, Vol124(3), p605-614" discloses a deep hole drilling depth optimization method based on a drilling force model. Methods The force of the drill pipe during the deep hole drilling process was analyzed. On this basis, an analytical model of the drilling axial force and torque changing with the drilling depth was established, and the model was predicted through the model coefficient calibration test. The drilling depth when the drill rod breaks during the deep hole drilling process optimizes the deep hole drilling depth. The method is based on the theoretical model and combined with the model coefficient calibration tes...

AI Technical Summary

Problems solved by technology

[0003] In order to overcome the shortcomings of poor practicability of existing deep hole processing depth optimization methods, the present invention provides a deep hole processing depth optimization method based on iterative learning

The present invention uses the deep hole processing cycle to correct the model coefficients in real time through an iterative learning method, continuously improves the prediction accuracy of the model, keeps it consistent with the real processing process, and solves the problem of insufficient prediction accuracy of the existing deep hole processing depth optimization method based on the analytical model problem, practical

Images