Optimal variable support size for mesh-free approaches using genetic algorithm

2021 ◽  
Vol 8 (4) ◽  
pp. 678-690 ◽  
Author(s):  
S. Hassouna ◽  
◽  
A. Timesli ◽  
Keyword(s):  
Genetic Algorithm ◽  
Local Density ◽  
Problem Complexity ◽  
Shape Functions ◽  
Mesh Free ◽  
Optimal Value ◽  
Approximation Errors ◽  
New Strategy ◽  
Support Size ◽  
Variable Support

The main difficulty of the meshless methods is related to the support of shape functions. These methods become stable when sufficiently large support is used. Rather larger support size leads to higher calculation costs and greatly degraded quality. The continuous adjustment of the support size to approximate the shape functions during the simulation can avoid this problem, but the choice of the support size relative to the local density is not a trivial problem. In the present work, we deal with finding a reasonable size of influence domain by using a genetic algorithm coupled with high order mesh-free algorithms which the optimal value depends on the accuracy and stability of the results. The proposed strategy provides guarantees about the growth of approximation errors, monitor the level of error, and adapt the evaluation strategy to reach the required level of accuracy. This allows the adaptation of the proposed algorithm with problem complexity. This new strategy in meshless approaches are tested on some examples of structural analysis.

Optimization of Cutting Parameters in Milling Process Using Genetic Algorithm and ANOVA (March 2020)

2020 ◽  
Vol 38 (10A) ◽  
pp. 1489-1503
Author(s):  
Marwa Q. Ibraheem
Keyword(s):  
Genetic Algorithm ◽  
Tool Wear ◽  
Material Removal Rate ◽  
Material Removal ◽  
Removal Rate ◽  
Cutting Speed ◽  
Cutting Parameters ◽  
Depth Of Cut ◽  
Optimal Value ◽  
Optimization Of Cutting Parameters

In this present work use a genetic algorithm for the selection of cutting conditions in milling operation such as cutting speed, feed and depth of cut to investigate the optimal value and the effects of it on the material removal rate and tool wear. The material selected for this work was Ti-6Al-4V Alloy using H13A carbide as a cutting tool. Two objective functions have been adopted gives minimum tool wear and maximum material removal rate that is simultaneously optimized. Finally, it does conclude from the results that the optimal value of cutting speed is (1992.601m/min), depth of cut is (1.55mm) and feed is (148.203mm/rev) for the present work.

Automatic Refinement of FE Shell Models Based on a Local Energy Function

2013 ◽  
Author(s):  
Antonio Carminelli ◽  
Giuseppe Catania
Keyword(s):  
Potential Energy ◽  
Local Density ◽  
Free Vibration Analysis ◽  
Element Model ◽  
Companion Paper ◽  
Total Potential Energy ◽  
Shape Functions ◽  
Total Potential ◽  
Error Functional ◽  
Local Patch

This paper deals with an adaptive refinement technique of a B-spline degenerate shell finite element model, for the free vibration analysis of curved thin and moderately thick-walled structures. The automatic refinement of the solution is based on an error functional related to the density of the total potential energy. The model refinement is generated by locally increasing, in a sub-domain R of a local patch domain, the number of shape functions while maintaining constant the functions polynomial order. The local refinement strategy is described in a companion paper, written by the same authors of this paper and presented in this Conference. A two-step iterative procedure is proposed. In the first step, one or more sub domains to be refined are identified by means of a point-wise error functional based on the system total potential energy local density. In the second step, the number of shape functions to be added is iteratively increased until the difference of the total potential energy, calculated on the sub domain between two iteration, is below a user defined tolerance. A numerical example is presented in order to test the proposed approach. Strengths and limits of the approach are critically discussed.

A new approach for indexing powder diffraction data based on whole-profile fitting and global optimization using a genetic algorithm

1999 ◽  
Vol 6 (2) ◽  
pp. 87-92 ◽  
Author(s):  
Benson M. Kariuki ◽  
Scott A. Belmonte ◽  
Malcolm I. McMahon ◽  
Roy L. Johnston ◽  
Kenneth D. M. Harris ◽  
...  
Keyword(s):  
Genetic Algorithm ◽  
Powder Diffraction ◽  
Diffraction Data ◽  
Lattice Parameters ◽  
Powder Diffraction Data ◽  
Diffraction Profile ◽  
New Approach ◽  
Profile Fitting ◽  
New Strategy ◽  
A New Technique

This paper describes a new technique for indexing powder diffraction data. The lattice parameters (unit-cell dimensions) {a,b,c,α,β,γ} define the parameter space of the problem, and the aim is to find the optimal lattice parameters for a given experimental powder diffraction pattern. Conventional methods for indexing consider the measured positions of a limited number of peak maxima, whereas this new approach considers the whole powder diffraction profile. This new strategy offers several advantages, which are discussed fully. In this approach, the quality of a given set of lattice parameters is determined from the profile R-factor, R wp, obtained following a Le Bail-type fit of the intensity profile. To find the correct lattice parameters (i.e. the global minimum in R wp), a genetic algorithm has been used to explore the R wp(a,b,c,α,β,γ) hypersurface. (Other methods for global minimization, such as Monte Carlo and simulated annealing, may also be effective.) Initially, a number of trial sets of lattice parameters are generated at random, and this `population' is then allowed to evolve subject to well defined evolutionary procedures for mating, mutation and natural selection (the fitness of each member of the population is determined from its value of R wp). Examples are presented to demonstrate the success and underline the potential of this new approach for indexing powder diffraction data.

A Boundary Enhancement for the Stabilized Conforming Nodal Integration of Galerkin Meshfree Methods

2015 ◽  
Vol 12 (02) ◽  
pp. 1550009 ◽  
Author(s):  
Dongdong Wang ◽  
Ming Sun ◽  
Pinkang Xie
Keyword(s):  
Transformation Method ◽  
Meshfree Methods ◽  
Shape Functions ◽  
Problem Domain ◽  
Nodal Integration ◽  
Support Size ◽  
Galerkin Meshfree ◽  
Integration Constraint ◽  
Linear Field ◽  
Stabilized Conforming Nodal Integration

The stabilized conforming nodal integration (SCNI) has been successfully developed for Galerkin meshfree methods based upon the linear exactness requirement. In this study, it is shown that for a given problem domain, when the support of the meshfree shape functions associated with the interior nodes do not cover the essential boundary, the linear exactness can be perfectly achieved by the standard SCNI formulation. On the other hand, when the essential boundary lies in the support of the meshfree shape functions of the interior nodes, a linear field may not be exactly obtained with the original SCNI formulation where the essential boundary conditions are enforced via the nodally exact transformation method, and the error even becomes more pronounced with the increase of support size. To resolve this issue, a flux term associated with the essential boundary is recovered in the variational formulation and it turns out to be proper to keep this term since the meshfree shape functions of interior nodes usually do not vanish on the boundary. Consequently the original SCNI integration constraint is revised and the stiffness matrix is enhanced by an additional stiffness contribution from the flux integration along the essential boundary. It is demonstrated that the proposed enhanced formulation is capable of exactly reproducing linear fields regardless of the support sizes. Moreover, several benchmark examples reveal that the present SCNI formulation with boundary enhancement yields better accuracy compared with the original SCNI approach, particularly for meshfree discretizations with larger support sizes.

scholarly journals Real-Coded Quantum-Inspired Genetic Algorithm-Based BP Neural Network Algorithm

2015 ◽  
Vol 2015 ◽  
pp. 1-10 ◽  
Author(s):  
Jianyong Liu ◽  
Huaixiao Wang ◽  
Yangyang Sun ◽  
Chengqun Fu ◽  
Jie Guo
Keyword(s):  
Neural Network ◽  
Genetic Algorithm ◽  
Bp Neural Network ◽  
Search Space ◽  
Descent Method ◽  
Gradient Descent Method ◽  
Quantum Genetic Algorithm ◽  
Binary Coding ◽  
Optimal Value ◽  
Neural Network Algorithm

The method that the real-coded quantum-inspired genetic algorithm (RQGA) used to optimize the weights and threshold of BP neural network is proposed to overcome the defect that the gradient descent method makes the algorithm easily fall into local optimal value in the learning process. Quantum genetic algorithm (QGA) is with good directional global optimization ability, but the conventional QGA is based on binary coding; the speed of calculation is reduced by the coding and decoding processes. So, RQGA is introduced to explore the search space, and the improved varied learning rate is adopted to train the BP neural network. Simulation test shows that the proposed algorithm is effective to rapidly converge to the solution conformed to constraint conditions.

scholarly journals Numerical solution of Drinfeld–Sokolov–Wilso system by using modified Adomian decomposition method

2021 ◽  
Vol 23 (1) ◽  
pp. 590
Author(s):  
Badran Jasim Salim ◽  
Oday Ahmed Jasim ◽  
Zeiad Yahya Ali
Keyword(s):  
Genetic Algorithm ◽  
Exact Solution ◽  
Differential Equations ◽  
Decomposition Method ◽  
Adomian Decomposition Method ◽  
Linear Partial Differential Equations ◽  
Modified Adomian Decomposition Method ◽  
Adomian Decomposition ◽  
Optimal Value ◽  
Mean Square Errors

<p class="Char">In this paper, the modified Adomian decomposition method (MADM) is usedto solve different types of differential equations, one of the numerical analysis methods for solving non linear partial differential equations (Drinfeld–Sokolov–Wilson system) and short (DSWS) that occur in shallow water flows. A Genetic Algorithm was used to find the optimal value for the parameter (a). We numerically solved the system (DSWS) and compared the result to the exact solution. When the value of it is low and close to zero, the MADM provides an excellent approximation to the exact solution. As well as the lower value of leads to the numerical algorithm of (MADM) approaching the real solution.  Finally, found the optimal value when a=-10 by using the Genetic Algorithm (G-MADM). All the computations were carried out with the aid of Maple 18 and Matlab to find the parameter value (a) by using the genetic algorithm as well as to figures drawing. The errors in this paper resulted from cut errors and mean square errors.</p>

Deploying massive runs of evolutionary algorithms with ECJ and Hadoop: Reducing interest points required for face recognition

2016 ◽  
Vol 32 (5) ◽  
pp. 706-720 ◽  
Author(s):  
Francisco Chávez ◽  
Francisco Fernández de Vega ◽  
Daniel Lanza ◽  
César Benavides ◽  
Juan Villegas ◽  
...  
Keyword(s):  
Genetic Algorithm ◽  
Big Data ◽  
Face Recognition ◽  
Evolutionary Algorithms ◽  
Real Life ◽  
Interest Points ◽  
Desktop Computer ◽  
New Strategy ◽  
Data Infrastructures ◽  
Mapreduce Model

In this paper we present a new strategy for deploying massive runs of evolutionary algorithms with the well-known Evolutionary Computation Library (ECJ) tool, which we combine with the MapReduce model so as to allow the deployment of computing intensive runs of evolutionary algorithms on big data infrastructures. Moreover, by addressing a hard real life problem, we show how the new strategy allows us to address problems that cannot be solved with more traditional approaches. Thus, this paper shows that by using the Hadoop framework ECJ users can, by means of a new parameter, choose where the run will be launched, whether in a Hadoop based infrastructure or in a desktop computer. Moreover, together with the performed tests we address the well-known face recognition problem with a new purpose: to allow a genetic algorithm to decide which are the more relevant interest points within the human face. Massive runs have allowed us to reduce the set from about 60 to just 20 points. In this way, recognition tasks based on the solution provided by the genetic algorithm will work significantly quicker in the future, given that just 20 points will be required. Therefore, two goals have been achieved: (a) to allow ECJ users to launch massive runs of evolutionary algorithms on big data infrastructures and also (b) to demonstrate the capabilities of the tool to successfully improve results regarding the problem of face recognition.

Reinforcement Design for Composite Laminate with Large Cutout by a Genetic Algorithm Method

2013 ◽  
Vol 631-632 ◽  
pp. 754-758
Author(s):  
Zhong Lei Sun ◽  
Mei Ying Zhao ◽  
Li Long Luo
Keyword(s):  
Genetic Algorithm ◽  
Optimization Method ◽  
Stacking Sequence ◽  
Sequence Optimization ◽  
Selection Probability ◽  
Stacking Sequence Optimization ◽  
Simple Genetic Algorithm ◽  
New Strategy ◽  
Genetic Algorithm Method ◽  
Reinforcement Design

A dual-zone reinforcement ply stacking sequence optimization method used for comosite laminate with large cutout is present. The optimization method utilized a new Genetic Algorithm. The new Genetic Algorithm introduced a new strategy which can improve the efficiency of the traditional Genetic Algorithm and overcome the shortages of the worse convergency and prematurity of the Simple Genetic Algorithm. In the new Genetic Algorithm, the selection probability and the mutation probability are self-adaptive. Compared with the Simple Genetic Algorithm, the new Genetic Algorithm method shows good consistency, fast convergency and practical feasibility. By using the new Genetic Algorithm, the reinforcement ply stacking sequence optimization method got reasonable symmetric and balance stacking sequence which could meet the design requirements.

scholarly journals Modelling and optimising a new hybrid ad-hoc network cooperation strategy performance using genetic algorithm

2021 ◽  
Vol 18 (2) ◽  
pp. 193-210
Author(s):  
Noor Jumaa ◽  
Abbas Allawy ◽  
Mustafa Shubbar
Keyword(s):  
Genetic Algorithm ◽  
Ad Hoc Network ◽  
Ad Hoc ◽  
Best Interests ◽  
Data Packets ◽  
Sufficient Power ◽  
Battery Capacity ◽  
New Strategy ◽  
Cooperation Strategy ◽  
Communication Source

The lifetime of an ad-hoc network depends on a mobile device?s limited battery capacity. In ad-hoc multi-hop communication, source nodes use intermediate nodes as a relay to communicate with remote destinations. As cooperation between nodes is restrained by their battery resources, it might not be in their best interests to always accept relay requests. Therefore, if all nodes decide how much energy to spend for relaying, selfish or non-cooperative nodes reduce cooperation by rejecting to forward packets to others, thereby leading to a dramatic drop in the network?s throughput. Three strategies have been founded to solve this problem: tit-for-tat, live-and-let-live, and selective drop. This research explored a new strategy in ad-hoc cooperation which resulted from the combination of the live-and-let-live and selective drop strategies. This new strategy is based on the suggestion to select fewer hops with a low drop percentage and sufficient power to stay alive after forwarding the data packets towards the destination or other relays at the route path. We used a genetic algorithm (GA) to optimise the cooperative problem. Moreover, the fitness equation of the GA population was designed according to the mixing of the two strategies, which resulted in a new optimized hybrid dynamic-static cooperation.

Sign in / Sign up

Export Citation Format

Share Document