Amarjeet Singh

How Improvements in Glowworm Swarm Optimization Can Solve Real-Life Problems

In order to solve real-life optimization problems, many Nature Inspired Optimization Techniques have come into existence over the last couple of years. Out of these, the categories of Swarm Intelligence Algorithms are gaining popularity due to their robustness and ease in applications. One such Swarm Intelligence algorithm is the Glowworm Swarm Optimization algorithm (GSO). This algorithm mimics the behavior of glowworms. The objective of this paper is to present a thorough survey of literature on various modifications and hybridizations of GSO along with their applications.

New variants of glowworm swarm optimization based on step size

In 2005, Krishnanand and Ghose (Multimodal function optimization using a glowworm metaphor with applications to collective robotics, 2005a), presented the idea of glowworm metaphor to determine multiple minima in the optimization problem arising in robotics applications. That research paper highlights the glowworm swarm behavior for determining multiple local minima for multimodal functions with application to robotics. Since then, a number of research papers have appeared to improve the performance of glowworm swarm optimization (GSO). In this paper, two major contributions are made. Firstly, a mathematical result is proved which shows that the step size of GSO has a significant influence on the convergence of GSO. Secondly, three variants of GSO are proposed which depend on different step size. Based on the implementation of the proposed variants and the original GSO on 15 benchmark problems, it is concluded that one of the proposed variants is a definite improvement over the original GSO and the remaining variants.

A "Never-Loose" Strategy to Play the Game of Tic-Tac-Toe

In much of the literature available to solve the tic-tac-toe board game, the common approaches, such as, co evolution, neural networks, evolutionary programming and genetic algorithm are used. In the present work we present a deterministic approach for playing tic-tac-toe game, in which 9 objective functions are defined to decide players best move. For choosing the best from the solutions generated, by combination of mutating zero as 1, certain axioms are defined. The beauty of this method lies in the fact that if the player decides a move with this method, he never loses the match whenever the player begins the game. We suspect that functions can be defined on the similar grounds for other existing board games and some of this work is in progress and will be reported elsewhere. Some implications on these lines have been made as recommendations in this paper.

Hybridized Gravitational Search Algorithms with Real Coded Genetic Algorithms for Integer and Mixed Integer Optimization Problems

In this paper, the Gravitational Search Algorithm (GSA) is hybridized with real coded Genetic Algorithm to solve Integer and Mixed Integer programming problems. The idea is based on two earlier papers of the authors. In the first paper, the authors proposed a methodology in which the Laplace Crossover and Power Mutation were embedded in Gravitational Search Algorithm and in the second paper, these algorithms were extended for the case of constrained optimization problems. In order to deal with integer variables, a special method is adopted. For dealing with the constraints the Deb’s technique is implemented. The original GSA and three new variants are tested on a set of benchmark problems available in literature. Based on the extensive numerical and graphical analysis of results it is concluded that one of the proposed variants outperform the original GSA and the other proposed variants.

Curve Fitting Using Gravitational Search Algorithm and Its Hybridized Variants

In many experimental studies in scientific applications a set of given data is to be approximated. This can be performed either by minimizing the least absolute deviation or by minimizing the least square error. The objective of this paper is to demonstrate the use of gravitational search algorithm and its recently proposed hybridized variants, called LXGSA, PMGSA and LXPMGSA, to fit polynomials of degree 1, 2, 3, or 4 to a set of N points. It is concluded that one of the hybridized version namely, LXPMGSA outperform all other variants for this problem.

Use of Evolutionary Algorithms to Play the Game of Checkers: Historical Developments, Challenges and Future Prospects

The objective of this paper is to study the historical development of computer programmers for playing the game of checkers. Since the game-playing is a NP-hard problem, it would be interesting to use evolutionary algorithms to solve them. The question is can a programme be developed which can beat humans with complete success, it may appears that some challenges may also be formed which may substantiate the argument of the paper. Further, these challenges also form a part of this study.

Reconstruction of 3D curves and surfaces using new variants of gravitational search algorithm

Abstract The problem of reconstruction of 3D curves and surfaces is modelled as a nonlinear optimization problem in which the objective function to be minimized is the error function between given data points and data points on the generated curve/surface. Although any optimization algorithm can be used to solve this problem, but due to the inherent complexities in the optimization model, heuristic algorithms are best suited. In this paper, a recently proposed nature motivated optimization technique, namely Gravitational Search Algorithm and its three variants are used to solve reconstruction problem. The approach is illustrated with the help of two examples- a helix and a dumbbell. The results are demonstrated and it is concluded that the approach is very promising in the area.

Novel hybridised variants of gravitational search algorithm for constraint optimisation

In an earlier paper, the authors proposed three new hybridised variants of gravitational search algorithm and real coded genetic algorithms for unconstrained optimisation problems, by hybridising GSA with Laplace crossover and power mutation. Experiments on a number of test problems, including CEC 2014 benchmarks, showed that the hybridised variant incorporating both Laplace crossover and power mutation emerged a winner in terms of efficiency and reliability by increased exploration and exploitation. This paper extends the hybridised variants proposed in the above paper, for the constrained optimisation making use of the Debs constraint handling mechanism. The performance of original GSA and the three proposed variants is investigated on a set of 24 constrained benchmark problems as given in CEC 2006. Based on a rigorous analysis of results, it is concluded that the variant hybridising GSA with Laplace crossover and power mutation outperforms all others.