|
|
 Âûïóñê 78
- Myachin A. Determination of centroids to increase the accuracy of ordinal-invariant pattern clustering
The work continues the research of constructing methods for analyzing patterns in parallel coordinates independent of the sequence of input data of the results. The basic operations on objects of ordinal-invariant pattern clusters are described. The assertion that the centroid of an ordinal-invariant pattern cluster belongs to the original cluster is proved, which allows one to estimate the intracluster object - centroid distances in the multidimensional feature space. Examples of revealing the structural similarity of objects in parallel coordinates are given. The main differences between the methods of analysis of patterns and cluster analysis are noted. The methodology of the centroid detection of the ordinal-invariant pattern-cluster is described. An algorithm for combining groups of objects based on their structural similarity, on the one hand, and minimizing intracluster distances, on the other, is proposed, which makes it possible to improve the accuracy of the final results and partially solve the problem of finding similar objects in the presence of error in the original data. The proposed algorithm uses the concept of intracluster distances “object - centroid” and satisfies the following conditions: endogenous determination of the number and composition of the desired groups of objects under study; low (relatively) computational complexity; independence of the original partition from the initial sequence of input data. The work of the proposed algorithm on classical data sets is demonstrated. The results of testing are presented and the clustering accuracy is increased.
- Ivanov D. Distribution of roles in coalitions of robots with limited communications based on the swarm interaction
The paper deals with the task of assigning roles to coalition robots with limited communications. The relevance of the use of robots coalitions, as well as the urgency of solving the role distribution problem in the coalitions of robots using the “one operator - one group of robots” scheme is shown. A formal formulation of the task of role distribution in the coalition of mobile robots is given. An analysis of existing approaches to the distribution of roles in groups of robots is given, such as solving the assignment problem by the Kuhn-Mankres algorithm, using the game theory apparatus, applying the methods of probability theory, the ant algorithm, and the method of propagating the control wave using a local conversion mechanism. An iterative approach to the distribution of roles in a group of robots, based on a decentralized management strategy and the principles of swarm interaction, is proposed. A method for the distribution of roles in coalitions of mobile robots and an algorithm for the actions of a single-party coalition robots implementing roles based on the proposed approach are described. The results of the study of the proposed approach, carried out with the help of computer simulation in coalitions of 100 robots in the distribution of three roles, are presented. The diagrams of roles distribution during iterations are given. The averaged results of computer modeling of the distribution of roles for different values ??of the radius of visibility of neighbors in the coalition are shown. The estimation of the error of the distribution of roles using the proposed algorithmically implemented method is made and a comparison with known approaches is carried out. The areas of possible practical application of the developed approach are shown.
- Maximov D. An optimal itinerary generation of large intellectual agent groups
Intelligence is represented in the Artificial General Intelligence (or Universal AI) approach as an information processor which consumes and gives out information. This information completely determines the system behavior. M. Hutter has obtained an optimal agent trajectory selecting in an abstract environment within this approach. However, the method demands number assessments of the reward in such a movement, and the way to obtain these assessments was not defined. The method does not also fit to estimate an agent group movement. In this paper, reward assessments are represented by the environment. The way to choose a trajectory for an agent group is also suggested. A group of intelligent agents which fulfill a set of tasks in parallel is represented by the tensor multiplication of corresponding processes in the Conway game category. The category corresponds to the environment and agents' moves. An optimal itinerary of the group is defined as a play with maximal total reward in the category. The reward is represented as a degree of certainty (visibility) of an agent goal, i.e., by some set, not a number. It is proved that such a reward determination may be used in the categorical construction of Conway games. The tensor product in the category is the operation of linear logic. Such logic is also used in selecting goals which the system can achieve of all visible ones. In the purpose, the whole goal set is represented as a lattice with the linear logic structure. The goal lattice is the truth value set of the logic in this case. The tensor product of the lattice elements corresponds to goals which are achieved in parallel. The system goals are chosen by the highest truth value of the lattice element which corresponds to they parallel achieving processes. Such a construction results in the formula of the optimal itinerary estimation in an abstract environment for an agent group.
- Glushchenko A. On efficiency of each of pi-controller parameters adjustment with neural tuner to reject dusturbances acting on heating furnaces
The scope of this research is to develop a modified version of previously proposed PI-controller neural tuner. Its rule base is improved. The new version is to be able to reject disturbances acting on heating furnaces at a steady-state mode. The following types of disturbances of limited amplitude are considered to be attenuated by the tuner– step and pulse disturbances acting the plant output signal, step and pulse disturbances acting the control action signal. The problem of appropriateness of KI parameter adjustment during the disturbance attenuation is considered for all mentioned above types of disturbances. Experiments are conducted using a muffle electroheating furnace SNOL 40/1200. Having analyzed the obtained results, the conclusion could be made that the most appropriate disturbance to adjust KI is the step one acting the control action signal, but the best course of action is to adjust this parameter for all types of disturbances. The tuner application allowed to reduce the disturbances rejection time by 18,8% ? 28,4% comparing to a conventional PI-controller.
- Kolokolov A., Lubinsky I., Yachno V. A method for constructing a frequecy analyzer of quasi-stationary signals
A method for frequency analysis of polyharmonic signals based on the time-frequency processing of the spectral pattern is proposed. The amplitude spectrum is determined in two steps in the proposed method. The amplitude spectrum of the analyzed signal is estimated by a comb of digital bandpass filters with highly overlapping frequency characteristics at the first stage.In order to find the amplitude spectrum, the output signals of the bandpass filters are rectified and smoothed by lowpass filters. At the second stage, the obtained spectrum is processed using a single-layer network of neural-like elements with delayed lateral inhibitory connections. The use of such a network allows to improve the characteristics of the frequency analyzer, namely, for a given accuracy and frequency resolution, to significantly reduce both the spectral envelope pulsations and the analysis time, which makes it possible to analyze shorter signals. The performance of the proposed method was tested with examples of harmonic and speech signal analysis for the case of using a comb of second-order digital bandpass filters with equal relative bandwidths and central frequencies evenly spaced on a logarithmic frequency scale. Such a choice of filter parameters of a frequency analyzer ensured equal accuracy of the analysis of the harmonic components of the signal under study.
- Zorine A., Kocheganov V. Statistical analysis and optimization of a tandem queuing systems under prolongable cyclic service
Tandem of controlling systems under prolongable cyclic service is presented. There are high and low-priority input flows in each system. Customers of the first system are serviced in class of cyclic algorithms: high-priority customers are serviced fixed amount of time and low-priority customers are then serviced another fixed amount of time. After service high-priority customers of the first system are transferred to the second one. In the second system, customers are serviced in the class of cyclic algorithms with prolongations: in addition to cyclic service it is possible to prolong service of high-priority customers in case amount of low-priority customers is below predefined threshold. Due to cybernetic approach it had become possible to build mathematical model, define necessary random variables and elements with specific distribution, which form Markov chain and represent problem setting. Also it had become possible to construct simulation model to conduct statistical experiments. Paper proposed an algorithm for determining when a stationary mode is reached. Estimators for system performance characteristics are also discussed. Stationary mode existence domain is investigated by means of experiments. In so doingt necessary conditions found by authors analytically are confirmed, and the possibility of their extension is demonstrated.
- Kiselev V. System of models for the assessment of programs of insurance of income in agriculture
The features of income insurance in agriculture are given. Unlike from the most popular multi-risk insurance programs the income insurance program insures from the losses due weather factors along with changes in forecast prices. Mathematical models of economy of the insurance company and agricultural firm with income insurance are given. These models differ from the models of crop insurance in the way of insurance payments describing. In addition to the random yield of insured crops the uncertain factors such as real prices at the harvest time are added in income insurance program. There are two types of information in the income insurance programs: information about the harvest and about the prices of products. The forecast price at the time of the insurance contract and the real price at the harvest time at which the insurance premiums calculated are in the calculations. Information on crop yields is presented by time series with different trends explained both by changes in weather conditions and by the of anthropogenic factors. A method for processing of such information is proposed. It represents the trend of time series as a linear spline where unknown coefficients are found by the least square method. Important information problems related to the forecast price for products are discussed. A method of using information both about the prices and cropness is proposed.
- Belov M., Novikov D. Models of a technology of complex activity control
The article continues the study of technology of complex activities [8]. The set of tasks of managing the development and mastering of new technologies of complex activity is considered. The problem of choosing standard solutions has been set and solved. In the framework of this task, it was shown that a uniform partition of the set of possible states of nature is “asymptotically” optimal from the point of view of minimizing the expected error, costs or / and entropy, and also maximizing the expected value of the level of learning and / or utility. For the tasks of resource distribution in aggregated technological networks, simple analytical algorithms for optimal resource allocation are proposed. The tasks of choosing the optimal strategy for switching from technology development to its productive use have been set and solved. For the case of a priori known nature characteristics, an optimal strategy was obtained and its properties were analytically investigated. For the case of unknown characteristics, a procedure is proposed that is optimal in the class of successive likelihood ratio rules. The results of simulation and analysis of approximations of the properties of the procedure are presented
- Gubiy E., Zorkaltsev V. Models and methods for reliability analysis of the energy supply of remote settlements
The mathematical models for the reliability analysis of energy supply of remote settlements are considered. The three-level complex of nested models is proposed. The lower level represents the model of functioning of the energy supply system of a remote settlement during a unit of time. The second level is a model of energy supply reliability analysis. This analysis is based on a multiple-fold imitation of the functioning of the energy supply system in randomly formed conditions. The values of energy demand and energy production, as well as the values of carryover energy reserves in storage devices are considered as random. The values of the demand and energy production for the imitations of functioning are formed by the Monte Carlo method from the given laws of probability of these quantities. The random value of the carryover energy reserves is formed by an algorithm that generates a Markov sequence of these reserves. The upper level represents the model for selecting the optimal composition of the means of ensuring reliability (power reserves in energy production and the capacity of energy storages). The mathematical expectation of the sum of the reduced costs for the operation of the energy supply system and the losses from the deficit is minimized. The values of such an objective function for a given means of ensuring reliability are determined as a result of a cycle of calculations on the model of reliability analysis. The results of studies of the reliability of biofuel supply from the energy plantation to a remote settlement, in the natural-climatic conditions of the coastal of the Baikal Lake are presented.
- Krygin A. Calculating and analytical methods of controlling the service of expanded engineering networks
The paper continues the study of the repair work optimization problem for engineering networks and is devoted to the application of these results for the practical assessment of the network section state and information support for decision making about extension of its service life. Three tasks were formulated for this: determining the optimal replacement period of the network's section, determining the integrated technical and economic indicator of the section status and selecting the best method for its diagnosis. These tasks were solved using the example of a heat supply network. During the study, it is became necessary to disclose in more detail some issues of repair optimization of engineering network sections, concerning section life cycle analysis and damage flow modeling in order to determine the dependency function of the total cost for maintenance from the time. It was shown that the damages flow on the heat pipeline section is described quite accurately by the Weibull distribution. The construction methods for dependency functions of the total number of damages and total maintenance costs from the time, using statistical data about section damages are suggested. À ñriteria for the best diagnostic method, a complex indicator, and the optimal period of a section replacing as a point of time at which the unit costs for its maintenance reach their minimum were constructed using this function.
|
|
