Intelligence Optimization Design Model of Deep Cement-Stirring Pile

Aiming at the phenomenon of the more conservative design of deep cement stirring pile currently, used optimization design theory such as genetic algorithm and particle swarm optimization, taken the cement consumption as the object function, taken replacement rate, water-cement ratio, pile diameter and pile length as the design variables, composite foundation bearing capacity and settlement as restrictive conditions, the optimal design models are established respectively based on genetic algorithm and particle swarm optimization. Case studies have shown that these two established models are effective. By comparison, the particle swarm optimization model is the more effective one.

Download Full-text

Airfoil Topology Optimization using Teaching-Learning based Optimization

International Journal of Applied Metaheuristic Computing ◽

10.4018/ijamc.2015010102 ◽

2015 ◽

Vol 6 (1) ◽

pp. 23-34

Author(s):

Dushhyanth Rajaram ◽

Himanshu Akhria ◽

S. N. Omkar

Keyword(s):

Genetic Algorithm ◽

Particle Swarm Optimization ◽

Particle Swarm ◽

Swarm Optimization ◽

Spline Curves ◽

Conceptual Design Phase ◽

Design Variables ◽

Heuristic Technique ◽

Teaching Learning Based Optimization ◽

Teaching Learning

This paper primarily deals with the optimization of airfoil topology using teaching-learning based optimization, a recently proposed heuristic technique, investigating performance in comparison to Genetic Algorithm and Particle Swarm Optimization. Airfoil parametrization and co-ordinate manipulations are accomplished using piecewise b-spline curves using thickness and camber for constraining the design space. The aimed objective of the exercise was easy computation, and incorporation of the scheme into the conceptual design phase of a low-reynolds number UAV for the SAE Aerodesign Competition. The 2D aerodynamic analyses and optimization routine are accomplished using the Xfoil code and MATLAB respectively. The effects of changing the number of design variables is presented. Also, the investigation shows better performance in the case of Teaching-Learning based optimization and Particle swarm optimization in comparison to Genetic Algorithm.

Download Full-text

Optimization Methods for Estimation of Missing Data

Computational Intelligence for Missing Data Imputation, Estimation, and Management ◽

10.4018/978-1-60566-336-4.ch010 ◽

2010 ◽

pp. 210-232

Author(s):

Tshilidzi Marwala

Keyword(s):

Genetic Algorithm ◽

Particle Swarm Optimization ◽

Simulated Annealing ◽

Missing Data ◽

Missing Values ◽

Particle Swarm ◽

Optimization Methods ◽

Hill Climbing ◽

Swarm Optimization ◽

Design Variables

This chapter presents various optimization methods to optimize the missing data error equation, which is made out of the autoassociative neural networks with missing values as design variables. The four optimization techniques that are used are: genetic algorithm, particle swarm optimization, hill climbing and simulated annealing. These optimization methods are tested on two datasets, namely, the beer taster dataset and the fault identification dataset. The results that are obtained are then compared. For these datasets, the results indicate that genetic algorithm approach produced the highest accuracy when compared to simulated annealing and particle swarm optimization. However, the results of these four optimization methods are the same order of magnitude while hill climbing produces the lowest accuracy.

Download Full-text

Performance Comparison of Particle Swarm Optimization and Genetic Algorithm in Rolling Fin-Tube Heat Exchanger Optimization Design

Heat Transfer: Volume 2 ◽

10.1115/ht2008-56213 ◽

2008 ◽

Cited By ~ 4

Author(s):

Wutao Han ◽

Linghong Tang ◽

Gongnan Xie ◽

Qiuwang Wang

Keyword(s):

Genetic Algorithm ◽

Particle Swarm Optimization ◽

Heat Exchanger ◽

Pressure Drop ◽

Optimization Design ◽

Particle Swarm ◽

Design Parameters ◽

Swarm Optimization ◽

Volume Weight ◽

Fin Tube

A method for optimization designs of rolling fin-tube heat exchangers was put forward with Particle Swarm Optimization (PSO) and Genetic Algorithm (GA), respectively. The length of tube bundles, the row numbers of tubes, the width of heat exchanger core and fin pitch were used as the optimization variables. The allowable pressure drop and heat exchange requirements were considered as restrictive conditions. According to specific design requirements, the volume, weight or pressure drop may be chosen as the optimization objective function. In the same design parameters, ranges of the search variables and restrictive conditions, optimization results compared with GA, the minimum volume, weight and pressure drop PSO could decrease by 3.34%, 4.31% and 14.04%, respectively, and corresponding CPU time could be reduced by 32.39%, 40.23% and 33.45%, respectively. In the fields of optimization designs of heat exchanger, Particle Swarm Optimization is a promising optimization method.

Download Full-text