Research And Application Of Multi-Objective Particle Swarm Optimization Algorithm

Multi-objective problems exist in request,to tackle these challenges,people have reasearched algorithm and applied it.One of them is particle swarm algorithm.We apply this algorithm because of its advantages of simplicity and fewer parameters.However,the algorithm also has shortcomings,such as the particles are trapped in a narrow space.The particle cannot scour the whole space,and the particle information exchangeability in the external archive is low.We made the following improvements to the algorithm and used it for the production process of ethylene:(1)The gene transfer method is added to the algorithm.We can think of the particle as a chromosome,so the position’s information on the particle can be thought of as the gene on the chromosome.The transfer method uses cutpaste and copy-paste operations to change the location of the particles in the external archive.This method strengthens particle information exchange,enriches the non-dominated solutions of external archives,improves the scope and the conjunction.(2)Quantum behavior is added to the algorithm,so as to promote the convergence speed of the algorithm.In the classical algorithm,the direction of a particle is affected by its velocity,while the particle in the quantum behavior has no vector.Instead,the attractor is obtained through the potential well,which can ensure that the probability P appears at any position in the space,which is good to for exploring of the algorithm.The dynamic explored strategy adjusts the exploring method of the algorithm through a dynamically changing parameter.This method can ensure that the algorithm will explore the further location at first.The particles will explore the deeper optimization according to the better location in the later stage,thus enriching the scope and the conjunction.(3)The improved algorithm was applied to the ethylene production process with the yield of ethylene and propylene as the target,so as to verify the results of the algorithm.The temperature of the pipe wall was taken as the operating variable,the amount of naphtha and the temperature of the nozzle were taken as the constraint conditions,and the model of the ethylene cracking was established by the equilibrium equation.After optimization,a decision scheme is provided to check the result of the algorithm.