Roman Šenkeřík

Application of Analytic Programming for Evolutionary Synthesis of Control Law—Introduction of Two Approaches

This research deals with an evolutionary synthesis of control law for Logistic equation, which is a discrete chaotic system. The novelty of the research is that an Analytic Programming (AP), which is a tool for symbolic regression, is used for the synthesis of feedback controller for chaotic system. This work introduces and compares two approaches representing blackbox type cost function, as well as not-blackbox type cost function. These two approaches are used for the purpose of stabilisation of the higher periodic orbits, which stand for oscillations between several values of chaotic system. The work consists of the descriptions of analytic programming as well as chaotic system and used cost functions. For experimentation, Self-Organising Migrating Algorithm (SOMA) and Differential Evolution (DE) were used.

Investigation on evolutionary predictive control of chemical reactor

This paper discusses predictive control of chemical reactor by means of evolutionary algorithm named SOMA - the Self-Organizing Migrating Algorithm, that can be classified as swarm intelligence or memetic algorithm. The SOMA algorithm was used for multiple input-multiple output control of reactor model after static optimization. The MIMO control was defined for 5 inputs and 2 outputs with total number of 12 unknown variables. This paper is based on simulation results from optimization of static parameters that has been used to set up reactor for its evolutionary control.

Steganography content detection by means of feedforward neural network

This paper describes the technique for revealing of steganography content in images by means of feedforward neural network. The work is connected with cryptography for information encoding and additional method for information hiding, which is called steganography. The substance of the paper explains the principle of hidden information detection in images like JPEG format by means of feedforward neural network. The results show successful detection of four stego techniques by means of artificial neural network with one hidden layer.

Swarm virus - Next-generation virus and antivirus paradigm?

Abstract In this article, we outline a possible dynamics, structure, and a behavior of a hypothetical (up to now) swarm malware as a background for a future antimalware system. We suggest how to capture and visualize behavior of such malware when it walks through the file system of an operating system. The swarm virus prototype, designed here, mimics a swarm system behavior and thus follows the main idea underlying the swarm intelligence algorithms. The information of the prototypes behavior is stored and visualized in the form of a complex network, reflecting virus communication and swarm behavior. The network nodes are then individual virus instances. The network has certain properties associated with its structure that can be used by the virus instances in its activities like locating target and executing payload on the right object. As the paper shows, the swarm behavior pattern can be incorporated also to an antimalware systems, and can be analyzed for a future computer system protection.

Complex Networks in Particle Swarm

This chapter presents an proposal of methodology for converting the inner dynamics of PSO algorithm into complex network. The motivation is in the recent trend of adaptive and learning methods for improving the performance of evolutionary computational techniques. It seems very likely that the complex network and its statistical characteristics can be used within those adaptive approaches. The network analysis also provides usefull insight into the inner dynamic of PSO. The methodology described in this chapter uses the communication in the swarm for construction of the network.

Gallery of Evolutionary Networks

This chapter is a graphical overview - a gallery of selected networks that have been obtained during our experiments. The gallery contains samples coming from different algorithms with attention on its beauty (as we hope) and shall serve as the visual motivation-inspiration for new experiments. Visualization of those networks has been done on different principles with only one aim: to show a variety of visualizations. Beside standard visualization, are present also visualizations like community, degree centralities, etc.

Randomization and Complex Networks for Meta-Heuristic Algorithms

This chapter deals with the hybridization of the chaos driven heuristics concept and complex networks framework for meta-heuristic. This research aims on the experimental investigations on the time development and influence of different randomization types, different strategies for Differential Evolution (DE) through the analysis of complex network as a record of population dynamics and indices selection. The population is visualized as an evolving complex network, which exhibits non-trivial features such as adjacency graph, centralities, clustering coefficient and other attributes showing efficiency of the network. Experiments were performed for different DE strategies, several different randomization types and simple test functions.

The Architecture of Software Interface for BCI System

The basic idea of Brain Computer Interface (BCI) is the connection of brain waves with an output device through some interface. Aim of this article is to clarify the potential utilization of complex EEG signal in BCI system. For this purpose, the architecture of the software interface was designed and tested. The main task of the interface is to transfer brain activity signal into commands of intelligent robot.

Recent advances in swarm and evolutionary computation-foreword

Swarm intelligence, algorithms as well as classical evolutionary algorithms are fairly active area of research in the domain of bioinspired algorithm. Its use can be found in the wide spectra of theoretical research as well as technological applications. In the last years it is observable that swarm and evolutionary algorithms are subject of hybridization with unconventional algorithms that improve its performance and/or help to analyze and better understanding of its dynamics, that can be very often chaotic or exhibiting various interesting patterns. All those mutual intersections thus have become a vitally important part of science and engineering at the theoretical as well as the practical level of research. The most interesting and applicable notions are, for example, chaos control and chaos synchronization related to secure communications, evolutionary algorithms and complex networks control, synthesis and analysis among others. Swarm and evolutionary algorithms are modern heuristic methods, based on natural processes that have wide applicability in many industrial problems. Recently, the study of swarm and evolutionary algorithms and its intersection with unconventional algorithms is focused not only along the traditional trends but also on the understanding and analyzing principles, with the new intention of mutual intersections of these interesting fields of research. This special issue discusses the proposed intersection of interesting fields of research, i.e. deterministic chaos, complex systems and swarm and evolutionary algorithms. Special issue will contain participations focused for example on how can be dynamic of chaotic systems and its “randomness” used inside swarm and evolutionary algorithms in order to increase its performance. Also participations discussing existence of various kinds of dynamics and behavior pattern inside swarm and evolutionary algorithms, analysis, visualization and its impact on performance are presented in this issue. Issue consists of 7 original research papers and one survey paper that summarize recent original research intersections of the fields as deterministic chaos, complex networks with bio-inspired methods.