The Study Of Multiobjective Multifactorial Operation Optimization For Continuous Annealing Production Process

Continuous annealing is an important process in strip production.Because of its complex production process,there are some coupling relationships between many environmental parameters and control variables.Therefore,it is difficult to achieve the global optimal effect by setting control variables through manual experience.The above reasons make the quality of strip steel fluctuate greatly,the energy consumption is too high,and so on.In view of the above problems,this thesis establishes two kinds of operation optimization models for continuous production process from the perspective of multifactorial optimization,taking iron and steel enterprises as the background and facing the production process of single coil and multiple coil respectively.Based on the characteristics of the problem and the model,the adaptive multifactorial evolutionary algorithm(AdaMFEA)and the adaptive multiobjective multifactorial differential evolutionary algorithm(AdaMOMFDE)are designed to solve these models,so as to improve enterprise benefits.This thesis includes the following contents.(1)A multifactorial operation optimization problem(MF-OOP)model for the continuous annealing production line(MF-OOP-CAPL)is established.This model is established to improve strip quality and reduce energy consumption.The task is to achieve an optimal parameter setting so that the two optimization objctives,which iteract but do not conflict with each other,can be optimizaed simultaneously.(2)A multiobjective multifactorial operation optimization problem(MOMF-OOP)model for the CAPL is established in this thesis.The MOMF-OOP-CAPL model is established to improve strip quality,reduce energy consumption and ensure unit productivity.This model can deal with multiple optimization tasks,each of which is a multiobjective operation optimization problems(MOOPs),at the same time.(3)To solve the MF-OOP model,the AdaMFEA is proposed.To imporve the robustness and search efficiency,multiple crossover operators are adopted in the algoirthm and an adaptive selection strategy for these operators is designed according to their search results.Meanwhile,an individual learning strategy based on backtracking linear search and quasi-Newton method is also proposed.The experimental results based on Benchmark problem show that AdaMFEA can effectively improve the efficiency of traditional multifactorial evolutionary algorithm.The experimental results based on practical industrial problems show that AdaMFEA can effectively solve the multifactorial operation optimization problem in continuous annealing production process.(4)To efficiently solve this problem,we propose an adaptive multiobjective multifactorial differential evolution algorithm(AdaMOMFDE),in which multiple mutation operators are adopted and an adaptive selection strategy for these operators is designed according to their search results so as to accelerate the convergence speed and improve the robustness of the algorithm.Experimental results on Benchmark problems show that the AdaMOMFDE that solves multiple optimization tasks simultaneously is better than the other advanced multiobjective evolutionary algorithms that only optimize one task at a time.Moreover,experimental results on practical problems illustrate that the AdaMOMFDE is effective and superior to the traditional multiobjective multifactorial evolutionary algorithm in the literature for the operation optimization of CAPL.