Multi-objective Optimization Of Process Industry Based On Global Fast Optimization Of CO Method

Production scheduling is an important part of the production management of process industry,its scientific directly affects the enterprise’s efficiency and competitiveness.Production scheduling of process industry often has multi objectives,thousands of design variables and constraints,so it is very necessary to introduce an effective decomposition method.Collaborative optimization(CO)is a typical multidisciplinary design optimization method,which can decompose complex large-scale models into system-level and multiple parallel sub-disciplines,and complex large-scale models is greatly reduced after decomposition.So it is important to study the multi-objective optimization problem of process industry based on CO method.The contents are as follows:(1)This paper proposes an improved collaborative optimization(ICO)method with fast global optimization to solve the problems that CO is sensitive to the choice of initial point and needs too many iterations and prone to converge on a local optimum.Firstly,a method of distinguishing the local and global optimal is proposed by disciplinary optimization information Secondly,at system level,the mixed relaxation factor(it is composed of dynamic relaxation factor and static relaxation factor)is used to relax the equality constraint,in which dynamic relaxation factor makes design variable converge to the extreme point quickly,static relaxation factor makes design variable jump out of the local optima.Finally,disciplinary level optimization objective function is composed of two parts,which are consistent objective function and sub discipline optimal objective function.Use standard CO method and ICO method to optimize the classical numerical case and the reducer optimization case.Compare with simulation results,it can be concluded that the ICO method can quickly obtain the global optimal solution and has a good robustness under the condition that the maximum constraint value is small.(2)To study multi-objective optimization problem(MOP)of process industry based on ICO,with the research background of beer mashing workshop.Considering constraints of equipment processing,capacity,distribution and material balance,the paper uses "process industry intelligent modeling software platform" to establish model of maximum capacity and heat energy wastage based on ICO,system-level objective function is composed of the weight sum method of the sub-disciplinary optimal objective function.Finally,genetic algorithm is used to solve the original MOP model and the MOP model based on ICO.The simulation results show the ICO method can simplify the model and quickly obtain the global optimal solution when solving this problem.The improvement of ICO is effective and feasible,and it has obvious advantages in the MOP problem application.