752 publications from this institution
Computational optimization forms an integrated part of modern computational science. Any good design should intend to achieve certain optimality, though optimal solutions are often difficult to find in practice since uncertainty and nonlinearity always present in almost all real-world problems. As resources, time and money are always limited, optimization becomes even more important in practice. This workshop on Computational Optimization, Modelling and Simulation (COMS 2010) at ICCS 2010 will summarize the latest developments of optimization and modelling and their applications in science, engineering and industry
The use of frequency-derivation information incorporated with model-based parameter estimation (MBPE) to predict the scattering from conducting and dielectric two-dimensional tragets over a frequency band is summarised. Instead of computing a method of moment solution using a point-by-point approach, a rational function model is used to approximate the current as a function of frequency. The model coefficients are computed using frequency-derivative information at one frequency within the band to improve efficiency of conventional moment method approaches. Results show that the model based approach gives excellent results over a limited frequency band, and is much more efficient than the converntional point-by-point approach.
An integrated part of modern design practice in both engineering and industry is simulation and optimization. Significant challenges still exist, though huge progress has been made in the last few decades. This 5th workshop on Computational Optimization, Modelling and Simulation (COMS 2014) at ICCS 2014 will summarize the latest developments of optimization and modelling and their applications in science, engineering and industry. This paper reviews the past developments, the stateof-the-art present and the future trends, while highlighting some challenging issues in these areas. It can be expected that future research should focus on the data intensive applications, approximations for computationally expensive methods, combinatorial optimization, and large-scale applications.
Editorial: Special issue on Metaheuristics in Artificial Intelligence By Guest Editors-in-Chief : Xin-She Yang,Suash Deb and Ping-Feng Pai
Gravitational and other anomalies seen repeatedly in connection with solar eclipses have led to speculation about a possible gravitational shielding effect as the cause. Here we show that an unusual phenomenon that occurs only during solar eclipses, rapid air mass movement for the bulk of the atmosphere above normal cloud levels, appears to be a sufficient explanation for both the magnitude and behavior of the anomaly previously reported in these pages.
The “firefly algorithm” (FFA) is a modern metaheuristic algorithm, inspired by the behavior of fireflies. This algorithm and its variants have been successfully applied to many continuous optimization problems. This work analyzes the performance of the FFA when solving combinatorial optimization problems. In order to improve the results, the original FFA is extended and improved for self-adaptation of control parameters, and thus more directly balancing between exploration and exploitation in the search process of fireflies. We use a new population model to increase the selection pressure, and the next generation selects only the fittest between a parent and an offspring population. As a result, the proposed memetic self-adaptive FFA (MSA-FFA) is compared with other well-known graph coloring algorithms such as Tabucol, the hybrid evolutionary algorithm, and an evolutionary algorithm with stepwise adaptation of weights. Various experiments have been conducted on a huge set of randomly generated graphs. The results of these experiments show that the results of the MSA-FFA are comparable with other tested algorithms.
We examined the influence of process uniformity on the device performance for the spin-orbit torque (SOT) magnetic random access memory (MRAM) devices. By optimizing the lithography process through changing the multi-energies exposure compensation and pretreatment in developing, we demonstrated 1.6 times sigma improvement of the critical dimensions (CD), leading to 49.5%, 54.2%, and 63.2% sigma% reduction for SOT channel resistance (R <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">b</inf> ), magnetic tunnel junction (MTJ) resistance (R <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">min</inf> ) and switching current (I <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">C</inf> ) respectively. These promising results will help to deliver SOT-MRAM to the mass manufacturing in the semiconductor industry.
No abstract is provided for this article.
Engineering optimization problems usually contain multifaceted performance requirements that can be riddled with unquantifiable specifications and incompatible performance objectives. Such problems typically possess competing design requirements which are very difficult – if not impossible – to quantify and capture at the time of model formulation. There are invariably unmodelled design issues, not apparent at the time of model construction, which can greatly impact the acceptability of the model’s solutions. Consequently, when solving many “real life” mathematical programming applications, it is generally preferable to formulate several quantifiably good alternatives that provide very different perspectives to the problem. These alternatives should possess near-optimal objective measures with respect to all known modelled objective(s), but be fundamentally different from each other in terms of the system structures characterized by their decision variables. This solution approach is referred to as modelling-to-generate-alternatives (MGA). This study demonstrates how the nature-inspired, Firefly Algorithm can be used to concurrently create multiple solution alternatives that both satisfy required system performance criteria and yet are maximally different in their decision spaces. This new co-evolutionary approach is very computationally efficient, since it permits the concurrent generation of multiple, good solution alternatives in a single computational run rather than the multiple implementations required in previous MGA procedures.
This chapter introduces the fundamentals of algorithms in the context of data mining, optimization, and machine learning, including the feasibility, constraints, optimality, Lagrange multipliers, KKT conditions, and gradient-based techniques.
In the mathematical modelling of compactional flow in porous media, the constitutive relation is typically modelled in terms of a nonlinear relationship between effective pressure and porosity, and compaction is essentially poroelastic. However, at depths deeper than 1 km where pressure is high, compaction becomes more akin to a viscous one. Two mathematical models of compaction in porous media are formulated and the noninear equations are then solved numerically. The essential features of numerical profiles of poroelastic and viscous compaction are thus compared with asymptotic solutions. Two distinguished styles of density-driven compaction in fast and slow compacting sediments are analysed and shown in this paper.
