INTELLIGENT MODELS FOR CONTROL OF JET HYDRO-PROCESSING OF ROLLED STEEL DEFECTS

Abstract

The structure of an intelligent control system for the hydro-processing of surface defects in rolled products is presented, which corresponds to the concept of the European program "Industry 4.0" in the direction of energy saving and total automation of processes.
 The considered local RBNN hydro-processing models are tuned using the Kaczmarz-Widrow-Hoff algorithm. The model for identifying the current task of the pressure of the processing fluid takes into account the relationship between the thickness of the processed defect, changes in the composition of the working fluid and, as a result, its abrasive properties. To form a given pressure in the corresponding jet, the following neural model estimates the duration and polarity of the voltage that controls the electric drive of the jet control valve in real time.
 The small-dimensional models’ conveyor is a promising solution for controlling waterjet processing of metals in such areas as rounding sharp edges; grinding and polishing complex surfaces; deburring and cleaning of welds; surface preparation for coating; removal of defects from the surface; high-precision waterjet cutting of metals. The solution will increase the speed of TP from 1.2 to 1.96 [m/s]. Taking into account the small dimensions of the models and the relatively low speed of the TP, a software application based on the proposed models can be implemented in the control system for the hydro-processing of rolled metal products in the form of an inexpensive IIOT microcontroller module.