Abstract

Many process synthesis and design problems in engineering are actually mixed integer nonlinear programming problems (MINLP), because they contain both continuous and integer variables. These problems are generally recognized to be complex and intractable by virtue of the combinatorial characteristic. In order to effectively solve process synthesis and design problems, a global particle swarm optimization (GPSO) algorithm is proposed in this paper. GPSO algorithm makes two improvements on original particle swarm optimization (PSO) algorithm: first, it introduces a global inertia weight, which is beneficial for improving its global searching capacity during the whole optimization process; second, it adopts a mutation operation with a small probability, which enables the GPSO algorithm to get rid of the local optimum easily. Simulation results show that the GPSO algorithm has high efficiency on finding the optimal solutions, and it has stronger convergence than the other four particle swarm optimization algorithms.

Highlights

  • I N order to establish an optimal construction, the selection, arrangement, and operation should be implemented for processing units, and this procedure is defined as process synthesis [1]

  • EXPERIMENTAL RESULTS AND ANALYSIS In order to verify the performance of the global particle swarm optimization algorithm (GPSO), eight unconstrained problems are selected, and they are given by f1(x) =

  • The population size P S is reset to 80. 50 runs are conducted in each case, and the optimization results are shown in Table 3: According to Table 3, all the five particle swarm optimization algorithm (PSO) algorithms can find the best solutions for all four process synthesis and design problems

Read more

Summary

INTRODUCTION

I N order to establish an optimal construction, the selection, arrangement, and operation should be implemented for processing units, and this procedure is defined as process synthesis [1]. The process synthesis and design problem belongs to a kind of constrained optimization problem. It has both integer and real variables, and is associated with some equality and inequality constraints. Vik,j and xki,j are the velocity component and position component of the ith particle at generation k, vik,+j 1 and xki,+j 1 are the veloc-. Chuanhu Chen et al.: Process synthesis and design problems based on a global particle swarm optimization algorithm ity component and position component of the ith particle at generation k + 1

A GLOBAL PARTICLE SWARM OPTIMIZATION ALGORITHM
EXPERIMENTAL RESULTS AND ANALYSIS
CONCLUSION
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call