Raka Jovanovic

An ant colony optimization algorithm with improved pheromone correction strategy for the minimum weight vertex cover problem

Abstract: The minimum weight vertex cover problem is an interesting and applicable NP-hard problem that has been investigated from many different aspects. The ant colony optimization metaheuristic is a relatively new technique that was successfully adjusted and applied to many hard combinatorial optimization problems, including the minimum weight vertex cover problem. Some kind of hybridization or exploitation of the knowledge about specific problem often greatly improves the performance of standard evolutionary algorithms. In this article we propose a pheromone correction heuristic strategy that uses information about the best-found solution to exclude suspicious elements from it. Elements are suspicious if they have some undesirable properties that make them unlikely members of the optimal solution. This hybridization improves pure ant colony optimization algorithm by avoiding early trapping in local convergence. We tested our algorithm on numerous test-cases that were used in the previous research of the same problem and our algorithm uniformly performed better, giving slightly better results in significantly shorter time.

Ant colony optimization algorithm with pheromone correction strategy for the minimum connected dominating set problem

In this paper an ant colony optimization (ACO) algorithm for the minimum connected dominating set problem (MCDSP) is presented. The MCDSP become increasingly important in recent years due to its applicability to the mobile ad hoc networks (MANETs) and sensor grids. We have implemented a one-step ACO algorithm based on a known simple greedy algorithm that has a significant drawback of being easily trapped in local optima. We have shown that by adding a pheromone correction strategy and dedicating special attention to the initial condition of the ACO algorithm this negative effect can be avoided. Using this approach it is possible to achieve good results without using the complex two-step ACO algorithm previously developed. We have tested our method on standard benchmark data and shown that it is competitive to the existing algorithms. [Projekat Ministarstva nauke Republike Srbije, br. III-44006]

A chain heuristic for the Blocks Relocation Problem

In the Blocks Relocation Problem (BRP) one is given a block retrieval sequence and is concerned with determining a relocation pattern minimizing the total number of moves required to enforce the given retrieval sequence. The importance of the BRP has been constantly growing in recent years, as a consequence of its close connection with the operations inside of a container terminal. Due to the complexity of the BRP, a large number of methods has been developed for finding near optimal solutions. These methods can be divided in two main categories greedy heuristics and more complex methods. The latter achieve results of higher quality, but at the cost of very long execution times. In many cases, this increased calculation time is not an option, and the fast heuristic methods need to be used. Greedy heuristic approaches, in general, apply the heuristic based only on the properties of the block that is being relocated and the current state of the bay. In this paper we propose a new heuristic approach in which when deciding where to relocate a block we also take into account the properties of the block that will be moved next. This idea is illustrated by improving the Min–Max heuristic for the BRP. We compare the new heuristic to several existing methods of this type, and show the effectiveness of our improvements. The tests have been conducted on a wide range of sizes of container bays, using standard test data sets.

A multi-heuristic approach for solving the pre-marshalling problem

Minimizing the number of reshuffling operations at maritime container terminals incorporates the pre-marshalling problem (PMP) as an important problem. Based on an analysis of existing solution approaches we develop new heuristics utilizing specific properties of problem instances of the PMP. We show that the heuristic performance is highly dependent on these properties. We introduce a new method that exploits a greedy heuristic of four stages, where for each of these stages several different heuristics may be applied. Instead of using randomization to improve the performance of the heuristic, we repetitively generate a number of solutions by using a combination of different heuristics for each stage. In doing so, only a small number of solutions is generated for which we intend that they do not have undesirable properties, contrary to the case when simple randomization is used. Our experiments show that such a deterministic algorithm significantly outperforms the original nondeterministic method. The improvement is twofold, both in the quality of found solutions, and in the computational effort.

Cuckoo search inspired hybridization of the Nelder-Mead simplex algorithm applied to optimization of photovoltaic cells

A new hybridization of the Cuckoo Search (CS) is developed and applied to optimize multi-cell solar systems; namely multi-junction and split spectrum cells. The new approach consists of combining the CS with the Nelder-Mead method. More precisely, instead of using single solutions as nests for the CS, we use the concept of a simplex which is used in the Nelder-Mead algorithm. This makes it possible to use the flip operation introduces in the Nelder-Mead algorithm instead of the Levy flight which is a standard part of the CS. In this way, the hybridized algorithm becomes more robust and less sensitive to parameter tuning which exists in CS. The goal of our work was to optimize the performance of multi-cell solar systems. Although the underlying problem consists of the minimization of a function of a relatively small number of parameters, the difficulty comes from the fact that the evaluation of the function is complex and only a small number of evaluations is possible. In our test, we show that the new method has a better performance when compared to similar but more compex hybridizations of Nelder-Mead algorithm using genetic algorithms or particle swarm optimization on standard benchmark functions. Finally, we show that the new method outperforms some standard meta-heuristics for the problem of interest.

A heuristic method for solving the problem of partitioning graphs with supply and demand

In this paper we present a greedy algorithm for solving the problem of the maximum partitioning of graphs with supply and demand (MPGSD). The goal of the method is to solve the MPGSD for large graphs in a reasonable time limit. This is done by using a two stage greedy algorithm, with two corresponding types of heuristics. The solutions acquired in this way are improved by applying a computationally inexpensive, hill climbing like, greedy correction procedure. In our numeric experiments we analyze different heuristic functions for each stage of the greedy algorithm, and show that their performance is highly dependent on the properties of the specific instance. Our tests show that by exploring a relatively small number of solutions generated by combining different heuristic functions, and applying the proposed correction procedure we can find solutions within only a few percent of the optimal ones.

An ant colony optimization algorithm for partitioning graphs with supply and demand

Graphical abstractDisplay Omitted HighlightsAnt colony optimization algorithm for the problem of partitioning graphs with supply and demand.Very effective method manages to find optimal solutions in more that 50% of the test instances.Average relative error of less than 0.5% when compared to known optimal solutions.Method analyzed on general graphs, Halin graphs, series-parallel graphs and trees. In this paper we focus on finding high quality solutions for the problem of maximum partitioning of graphs with supply and demand (MPGSD). There is a growing interest for the MPGSD due to its close connection to problems appearing in the field of electrical distribution systems, especially for the optimization of self-adequacy of interconnected microgrids. We propose an ant colony optimization algorithm for the problem. With the goal of further improving the algorithm we combine it with a previously developed correction procedure. In our computational experiments we evaluate the performance of the proposed algorithm on trees, 3-connected graphs, series-parallel graphs and general graphs. The tests show that the method manages to find optimal solutions for more than 50% of the problem instances, and has an average relative error of less than 0.5% when compared to known optimal solutions.

Gaussian-induced rotation in periodic photonic lattices

We numerically investigate time-dependent rotation of counterpropagating mutually incoherent self-trapped Gaussian beams in periodic optically induced fixed photonic lattices. We demonstrate the relation between such rotation and less confined discrete solitonic solutions.

A greedy method for optimizing the self-adequacy of microgrids presented as partitioning of graphs with supply and demand

In this paper we focus on solving the problem of maximal partitioning of graphs with supply and demand. The interest for this problem is due to the fact that it well represents the optimization of self-adequacy of interconnected microgrids. The chosen method is a heuristic based greedy algorithm which can be applied to very large problem instances in feasible time, which best relates to potential real life applications. To decrease the computational time of the algorithm, suitable auxiliary structures are introduced. To get the best performance, in the sense of finding high quality solutions, we have analyzed the behavior of the algorithm for three different heuristics. In the conducted experiments it has been show that the proposed method is very efficient in the case of small problem instances, for which it frequently manages to find optimal solutions. The performed test have demonstrated that the proposed method generally acquires solutions within 5-10% of the known optimal one.

An algorithm for automated segmentation for bleeding detection in endoscopic images

Wireless capsule endoscopy is an important advanced diagnostics method. It produces huge amount of images during travel through patients digestive tract and that usually requires automated analysis. One of the most important abnormalities is bleeding and automated segmentation for bleeding detection is an active research topic. In this paper we propose an algorithm for automated segmentation for bleeding detection in capsule endoscopy images. The algorithm uses block based segmentation where average saturation from the HSI model and skewness and kurtosis of uniform local binary patterns histogram are used as features for the support vector machine classifier. Support vector machine parameters are tuned using grid search. The proposed method was tested using standard benchmark images and compared with other approaches from literature using Dice similarity coefficient and misclassification error as metrics, where it obtained better results using simpler features.