30 October 2016 in Eastern-European Journal of Enterprise Technologies
Today in Ukraine there is an increased level of natural and technogenic threats and fire hazards. Therefore, an important task in identifying and assessing risks and threats is to determine fixed and variable factors that affect the potential for fires and to classify them by the available features. To solve the problem of classifying numerous factors of fires, we suggest using the method of building decision trees, which is a method of presenting rules in a hierarchical consistent structure where each object corresponds to a single node through which the decision is made. The use of the C4.5 algorithm helps build a branched decision tree and classify factors of fire danger. Three main classes of permanent environmental factors have been distinguished, which include land cover, topography, and climatic resources; the variable factors are the indices NDVI, DMP, and SWI. They, in turn, are divided into subclasses. The calculated weights can be used for simulating a fire hazard. The obtained values range from 0 to 1, where a value of 0 prevents natural fires (e. g., water surfaces), but values close to 1 indicate a high hazard potential of natural fires.The decision trees, obtained in the process of classification, are important for planning measures to prevent natural fires. They can also be used for zoning in terms of fire hazards in spatial modeling of fires, mathematical modeling of their effects, as well as in further monitoring and prediction of natural fires.
1 January 2021 in Advanced Information Technology
This paper is devoted to the problem of optimization of a function in -dimensional space, which, in general case, is polyextreme and undifferentiated. The new method of deformed stars in n-dimensional space was proposed. It is built on the ideas and principles of the evolutionary paradigm. Method of deformed stars is based on the assumption of using potential solutions groups. There by it allows to increase the rate of the accuracy and the convergence of the achieved result. Populations of potential solutions are used to optimize the multivariable function. In contrast to the classical method of deformed stars, we obtained a method that solves problems in -dimensional space, where the population of solutions consists of 3-, 4-, and 5-point groups. The advantages of the developed method over genetic algorithm, differential evolution and evolutionary strategy as the most typical evolutionary algorithms are shown. Also, experiments were performed to investigate the best configuration of method of deformed stars parameters.
1 January 2020 in Information systems and technologies security
In this paper are being considered the aspects of two variables function optimization problem solving, which, in general, is poly-extremal and undifferentiated. The classic methods of continuous optimization are not applicable in this case. One of the most commonly used methods of solving this problem is evolutionary algorithms, which can be divided into two classes. The first class includes algorithms where a potential offspring-solution is generated by two parent-solutions solutions, in the second case, the offspring-solution is generated by one parent-solution. There is deformed star method proposed where the population of parental solutions is 3, 4, and 5 point groups. The application of proposed method is shown to solve the optimization problem of fire monitoring system for buildings, which minimizes the time of its operation. The buildings where fire load can be both permanent and variable are considered. Such buildings include concert halls, nightclubs, supermarkets, logistics facilities and more. Fires at such buildings result in human sacrifice and serious material loss. Timely activation of the fire alarm system have great importance. The objective function of the problem is determined by the distance from the horizontal projections of the detectors to the sources of fire and the probability of triggering the detectors. The solution is optimizing location of fire detectors, taking into account their number and the fire load of the room. The advantages of the developed method over genetic algorithms, evolutionary strategies and differential evolution as the most typical evolutionary algorithms are shown. Numerical experiments were carried out, which showed the increased accuracy of calculations and the increased speed of method convergence.