• Journal of Practical Engineering Education
• /
• v.14 no.2
• /
• pp.305-312
• /
• 2022

In this paper, we focused on training to create and optimize a basic data prediction model. And we proposed a gradient descent training method of machine learning that is widely used to optimize data prediction models. It visually shows the entire operation process of gradient descent used in the process of optimizing parameter values required for data prediction models by applying the differential method and teaches the effective use of mathematical differentiation in machine learning. In order to visually explain the entire operation process of gradient descent, we implement gradient descent SW in a spreadsheet. In this paper, first, a two-variable gradient descent training method is presented, and the accuracy of the two-variable data prediction model is verified by comparison with the error least squares method. Second, a three-variable gradient descent training method is presented and the accuracy of a three-variable data prediction model is verified. Afterwards, the direction of the optimization practice for gradient descent was presented, and the educational effect of the proposed gradient descent method was analyzed through the results of satisfaction with education for non-majors.

• Journal of the Korea Convergence Society
• /
• v.5 no.4
• /
• pp.163-169
• /
• 2014

A stable neural network control scheme for unknown non-linear systems is developed in this paper. While the control variable is optimised to minimize the performance index, convergence of the index is guaranteed asymptotically stable by a Lyapnov control law. The optimization is achieved using a gradient descent searching algorithm and is consequently slow. A fast convergence algorithm using an adaptive learning rate is employed to speed up the convergence. Application of the stable control to a single input single output (SISO) non-linear system is simulated. The satisfactory control performance is obtained.

• 한국전산유체공학회:학술대회논문집
• /
• 2006.05a
• /
• pp.263-267
• /
• 2006

The path of a flapping airfoil during upstroke and down-stroke is optimized for maximum thrust and propulsive efficiency. The periodic flapping motion in combined pitch and plunge is described using Non-Uniform B-Splines(NURBS). A gradient based algorithm is employed for optimization of the NURBS parameters. Unsteady, low speed laminar flows are computed using a Navier-Stokes solver in a parallel computing environment based on domain decomposition. It is shown that the thrust generation is significantly improved in comparison to the sinusoidal flapping motion. For a high thrust generation, the airfoil stays at a high effective angle of attack for short durations.

• Steel and Composite Structures
• /
• v.21 no.6
• /
• pp.1389-1410
• /
• 2016

This paper proposes a hybrid of topological derivative-based level set method (LSM) and isogeometric analysis (IGA) for structural topology optimization. In topology optimization a significant drawback of the conventional LSM is that it cannot create new holes in the design domain. In this study, the topological derivative approach is used to create new holes in appropriate places of the design domain, and alleviate the strong dependency of the optimal topology on the initial design. Furthermore, the values of the gradient vector in Hamilton-Jacobi equation in the conventional LSM are replaced with a Delta function. In the topology optimization procedure IGA based on Non-Uniform Rational B-Spline (NURBS) functions is utilized to overcome the drawbacks in the conventional finite element method (FEM) based topology optimization approaches. Several numerical examples are provided to confirm the computational efficiency and robustness of the proposed method in comparison with derivative-based LSM and FEM.

• Coupled systems mechanics
• /
• v.13 no.1
• /
• pp.1-19
• /
• 2024

The effectiveness of coupling with a bolted rim is assessed in this research using a newly designed optimization algorithm. The current study, which is provided here, evaluates 10 contemporary metaheuristic approaches for enhancing the coupling with bolted rim design problem. The algorithms used are particle swarm optimization (PSO), crow search algorithm (CSA), enhanced honeybee mating optimization (EHBMO), Harmony search algorithm (HSA), Krill heard algorithm (KHA), Pattern search algorithm (PSA), Charged system search algorithm (CSSA), Salp swarm algorithm (SSA), Big bang big crunch optimization (B-BBBCO), Gradient based Algorithm (GBA). The contribution of the paper isto optimize the coupling with bolted rim problem by comparing these 10 algorithms and to find which algorithm gives the best optimized result. These algorithm's performance is evaluated statistically and subjectively.

• 한국전산유체공학회:학술대회논문집
• /
• 2005.04a
• /
• pp.191-196
• /
• 2005

For the improvement of aerodynamic performance of the turbine blade in a turbopump for the liquid rocket engine, the optimization of turbine profile shape has been studied. The turbine in a turbopump in this study is a partial admission of impulse type, which has twelve nozzles and supersonic inflow. Due to the separated nozzles and supersonic expansion, the flow field becomes complicates and shows oblique shocks and flow separation. To increase the blade power, redesign of the blade shape using CFD and optimization method was attempted. The turbine cascade shape was represented by four design parameters. For optimization, genetic algorithm based upon non-gradient search has been selected as a optimizer. As a result, the final blade has about 4 percent more blade power than the initial shape.

• Journal of computational fluids engineering
• /
• v.10 no.2
• /
• pp.54-59
• /
• 2005

For the improvement of aerodynamic performance of the turbine blade in a turbopump for the liquid rocket engine, the optimization of turbine profile shape has been studied. The turbine in a turbopump in this study is a partial admission of impulse type, which has twelve nozzles and supersonic inflow. Due to the separated nozzles and supersonic expansion, the flow field becomes complicate and shows oblique shocks and flow separation. To increase the blade power, redesign ol the blade shape using CFD and optimization methods was attempted. The turbine cascade shape was represented by four design parameters. For optimization, a genetic algorithm based upon non-gradient search hue been selected as an optimizer. As a result, the final blade has about 4 percent more blade power than the initial shape.

• Journal of the Society of Naval Architects of Korea
• /
• v.45 no.6
• /
• pp.590-600
• /
• 2008

This paper concerns the development of a designer friendly hull form parameterization and its coupling with advanced global optimization algorithms. As optimization algorithms, we choose the Partial Swarm Optimization(PSO) recently introduced to solve global optimization problems. Most general-purpose optimization softwares used in industrial applications use gradient-based algorithms, mainly due to their convergence properties and computational efficiency when a relatively few number of variables are considered. However, local optimizers have difficulties with local minima and non-connected feasible regions. Because of the increase of computer power and of the development of efficient Global Optimization (GO) methods, in recent years nongradient-based algorithms have attracted much attention. Furthermore, GO methods provide several advantages over local approaches. In the paper, the derivative-based SQP and the GO approach PSO are compared with their relative performances in solving some typical ship design optimization problem focusing on their effectiveness and efficiency.

• Journal of Institute of Control, Robotics and Systems
• /
• v.13 no.8
• /
• pp.790-797
• /
• 2007

This paper proposes a new measure of cost function ruggedness in local optimization with DEAS. DEAS is a computational optimization method developed since 2002 and has been applied to various engineering fields with success. Since DEAS is a recent optimization method which is rarely introduced in Korean, this paper first provides a brief overview and description of DEAS. In minimizing cost function with this non-gradient method, information on function shape measured automatically will enhance search capability. Considering the search strategies of DEAS are well designed with binary matrix structures, analysis of search behaviors will produce beneficial shape information. This paper deals with a simple quadratic function contained with various magnitudes of noise, and DEAS finds local minimum yielding ruggedness measure of given cost function. The proposed shape information will be directly used in improving DEAS performance in future work.

• Proceedings of the Korean Society of Medical Physics Conference
• /
• 2005.04a
• /
• pp.79-82
• /
• 2005

In this study, we have tested the feasibility of the convex non-linear objective model and the line search optimization method for the fluence map optimization (FMO). We've created the convex nonlinear objective function with simple bound constraints and attained the optimal solution using well-known gradient algorithms with an Armijo line search that requires sufficient objective function decrease. The algorithms were applied to 10 head-and-neck cases. The numbers of beamlets were between 900 and 2,100 with a 3 mm isotropic dose grid. Nonlinear optimization methods could efficiently solve the IMRT FMO problem in well under a minute with high quality for clinical cases.