Management of the hydraulic descaling from rolled sheet, based on genetic algorithms

Abstract

Hydraulic descaling is one of the most progressive method for scale removal in metallurgy. The most rational way to identify key reservoir parameters for hydraulic descaling is to provide the necessary liquid jets energy to remove the scale. These parameters are specific for each grade of steel. Specific scale deleting energy e * achieving is needed to provide the effectively scale removing. Scale density is defined by real time thermal images for each section of steel sheet. Value e * exceeding is due to the additional water flow. Also, we need to minimize the cost of management. Therefore the control criterion can be represented as function of three variables f(H,p,γ). Presented in the above problem belongs to a class of nonlinear optimization problems. That’s why it is advisable to use genetic algorithms. The system uses an integer encoding that match to the desired control variables value: the water pressure in the nozzles, beam position, angle nozzle. The authors have developed a fitness function of the system. The following statements have been used in algorithm : selection of roulette; single-point crossover; simple mutation. Authors have developed software to perform the calculation of the control variables values