Volume 20, Nº 2 (2021)

In modern conditions, in the field of the creation and use of existing and advanced space vehicles (SV), the issues of autonomy and survivability acquire particular relevance in the development and operation of small-mass spacecraft (SMS) for Earth remote sensing (ERS). The specificity of the small spacecraft lies in the fact that it is difficult to directly apply to the process of their creation the standard practice of using the system for ensuring the reliability of the rocket and space industry due to the lack of the ability to provide full structural redundancy of its onboard systems (OBS) associated with mass-dimensional and other restrictions. In this case, the tasks of developing model-algorithmic methods and approaches to ensuring the required level of indicators of structural reliability, survivability and, in general, the effectiveness of the functioning of the MCA OBS become of particular relevance. The problem of increasing the level of indicators of autonomy, survivability, efficiency of functioning of complex technical objects (CTO), which, in particular, SMS belong, is considered in the scientific literature in conjunction with solving problems of control, assessment and technical diagnostics of the state of the CTO reconfiguration (structural, functional, structural-functional reconfiguration) of CTO structures, management of its reserves, alternative and multi-mode control, analysis of fault tolerance and disaster recovery of CTO. However, all of these studies are fragmented, both at the methodological and methodological and technological levels. The article provides a generalized description of the combined methods and algorithms developed by the authors for solving the problems of synthesis of technologies and programs for controlling the OS reconfiguration to increase the survivability of the SMS. At the same time, these tasks are solved not in isolation, but in a comprehensive manner within the framework of the general problem of proactive management of the structural dynamics of SMS with or without the use of GCC tools, which ensures the efficiency, validity, completeness, isolation and consistency of synthesized management decisions. The novelty of the approach proposed in the article is that its authors, based on the concepts of integrated (system) modeling, proactive control of the structural dynamics of the OS SMS, as well as the intellectualization of the processes of proactive control of the OS SMS, developed methods and algorithms for the synthesis of technologies and programs. Control of the reconfiguration of the MCS BS, providing, firstly, the situational choice of the optimal sequence of operations and the allocation of SMS resources with and without the use of GCC facilities, and, secondly, effective parrying not only of the calculated ones, but also off-design emergency flight situations (EFS), as well as the operational restoration of the operability of its OS. The constructiveness of the proposed approach is illustrated by the example of solving the problem of flexible redistribution of information processing tasks between the OS SMS and the SMS GCC.

The paper proposes a method for fusioning multi-angle images implementing the algorithm for quasi-optimal clustering of pixels to the original images of the land surface. The original multi-angle images formed by the onboard equipment of multi-positional location systems are docked into a single composite image and, using a high-speed algorithm for quasi-optimal pixel clustering, are reduced to several colors while maintaining characteristic boundaries. A feature of the algorithm of quasi-optimal pixel clustering is the generation of a series of partitions with gradually increasing detail due to a variable number of clusters. This feature allows you to choose an appropriate partition of a pair of docked images from the generated series. The search for reference points of the isolated contours is performed on a pair of images from the selected partition of the docked image. A functional transformation is determined for these points. And after it has been applied to the original images, the degree of correlation of the fused image is estimated. Both the position of the reference points of the contour and the desired functional transformation itself are refined until the evaluation of the fusion quality is acceptable. The type of functional transformation is selected according to the images reduced in color, which later is applied to the original images. This process is repeated for clustered images with greater detail in the event that the assessment of the fusion quality is not acceptable. The purpose of present study is to develop a method that allows synthesizing fused image of the land surface from heteromorphic and heterogeneous images. The paper presents the following features of the fusing method. The first feature is the processing of a single composite image from a pair of docked source images by the pixel clustering algorithm, what makes it possible to isolate the same areas in its different parts in a similar way. The second feature consists in determining the functional transformation by the isolated reference points of the contour on the processed pair of clustered images, which is later applied to the original images to combine them. The paper presents the results on the synthesis of a fused image both from homogeneous (optical) images and from heterogeneous (radar and optical) images. A distinctive feature of the developed method is to improve the quality of synthesis, increase the accuracy and information content of the final fused image of the land surface.

The problem of stability analysis and its components of reliability and survivability is quite popular both in the field of telecommunications and in other industries involved in the development and operation of complex networks. The most suitable network model for this type of problem is a model that uses the postulates of graph theory. At the same time, the assumption of the random nature of failures of individual links of the telecommunications network allows it to be considered in the form of a generalized Erdos–Renyi model. It is well known that the probability of failure of elements can be interpreted in the form of a readiness coefficient and an operational readiness coefficient, as well as in the form of other indicators that characterize the performance of elements of a telecommunications network. Most approaches consider only the case of bipolar connectivity, when it is necessary to ensure the interaction of two end destinations. In modern telecommunications networks, services such as virtual private networks come to the fore, for which multipoint connections are organized that do not fit into the concept of bipolar connectivity. In this regard, we propose to extend this approach to the analysis of multi-pole and all-pole connections. The approach for two-pole connectivity is based on a method that uses the connectivity matrix as a basis, and, in fact, assumes a sequential search of all combinations of vertex sections, starting from the source and drain. This method leads to the inclusion of non-minimal cross-sections in the general composition, which required the introduction of an additional procedure for checking the added cross-section for non-excess. The approach for all-pole connectivity is based on a method that uses the connectivity matrix as a basis, and, in fact, assumes a sequential search of all combinations of vertex sections, not including one of the vertices considered terminal. A simpler solution was to control the added section for uniqueness. The approach for multipolar connectivity is similar to that used in the formation of the set of minimal all-pole sections and differs only in the procedure for selecting the combinations used to form the cross-section matrix, of which only those containing pole vertices are preserved. As a test communication network, the Rostelecom backbone network is used, deployed to form flows in the direction of "Europe-Asia". It is shown that multipolar sections are the most general concept with respect to two-pole and all-pole sections. despite the possibility of such a generalization, in practical applications it is advisable to consider particular cases due to their lower computational complexity.

In this article the task of determining the current position of pneumatic actuators is considered. The solution to the given task is achieved by using a technical vision system that allows to apply the fuzzy clustering method to determine in real time the center coordinates and the displacement position of a color label located on the mechatronic complex actuators. The objective of this work is to improve the accuracy of the moving actuator’s of mechatronic complex by improving the accuracy of the color label recognition. The intellectualization of process of the color shade recognition is based on fuzzy clustering. First, a fuzzy model is built, that allows depending on the input parameters of the color intensity for each of the RGB channels and the color tone component, to select a certain color in the image. After that, the color image is binarized and noise is suppressed. The authors used two defuzzification models during simulation a fuzzy system: one is based on the center of gravity method (CoG) and the other is based on the method of area ratio (MAR). The model is implemented based on the method of area ratio and allows to remove the dead zones that are present in the center of gravity model. The method of area ratio determines the location of the color label in the image frame. Subsequently, when the actuator is moved longitudinally, the vision system determines the location of the color label in the new frame. The color label position offset between the source and target images allows to determine the moved distance of the color label. In order to study  how noise affects recognition accuracy, the following digital filters were used: median, Gaussian, matrix and binomial. Analysis of the accuracy of these filters showed that the best result was obtained when using a Gaussian filter. The estimation was based on the signal-to-noise coefficient. The mathematical models of fuzzy clustering of color label recognition were simulated in the Matlab/Simulink environment. Experimental studies of technical vision system performance with the proposed fuzzy clustering model were carried out on a pneumatic mechatronic complex that performs processing, moving and storing of details. During the experiments, a color label was placed on the cylinder, after which the cylinder moved along the guides in the longitudinal direction. During the movement, video recording and image recognition were performed. To determine the accuracy of color label recognition, the PSNR and RMSE coefficients were calculated which were equal 38.21 and 3.14, respectively. The accuracy of determining the displacement based on the developed model for recognizing color labels was equal 99.7%. The defuzzifier speed has increased to 590 ns.

The paper proposes a hybrid method for training convolutional neural networks. The method consists of combining second and first-order methods for different elements of the architecture of a convolutional neural network. The hybrid convolution neural network training method allows to achieve significantly better convergence compared to Adam; however, it requires fewer computational operations to implement. Using the proposed method, it is possible to train networks on which learning paralysis occurs when using first-order methods. Moreover, the proposed method could adjust its computational complexity to the hardware on which the computation is performed; at the same time, the hybrid method allows using the mini-packet learning approach. The analysis of the ratio of computations between convolutional neural networks and fully connected artificial neural networks is presented. The mathematical apparatus of error optimization of artificial neural networks is considered, including the method of backpropagation of the error, the Levenberg-Marquardt algorithm. The main limitations of these methods that arise when training a convolutional neural network are analyzed. The analysis of the stability of the proposed method when the initialization parameters are changed. The results of the applicability of the method in various problems are presented.