• Energy Research

In the present article we will consider a class of associative machines with dynamic structure where the entrance signal exerts direct impact on the mechanism of association of output signals of experts. At the same time we are interested in such group of expert decisions at which separate expert responses unite not linearly through hierarchically organized lock networks. Hierarchical mixture of opinions of experts, along with simple mixture are examples of modular networks: neural network of a module if the calculations executed by it can be distributed on several subsystems processing different entrance signals and not crossed in the work. Output signals of these subsystems unite the integrative module which exit does not possess feedback with subsystems. In fact, the integrative module makes the decision as output signals of subsystems are grouped in the general output signal of system, and identifies what examples are samples for training of concrete modules. The most general definition of modular neural network: any set of algorithms of data processing, including algorithms of the artificial neural networks grouped for the solution of some uniform task. Automatically determine the class of associative machines with dynamic structure where the entrance signal exerts direct impact on the mechanism of association of output signals of experts, at the same time group of expert decisions at which separate expert responses unite not linearly through hierarchically organized lock networks is considered.

The paper deals with the management of multi-connected objects with multiple inputs and outputs, the so-called MIMO (multi-input multi-output) systems. Special attention is paid to the adaptation of SISO (single-input single-output) technologies to control MIMO systems. For this, it is proposed to carry out dynamic decoupling of control channels using a pre-switched on cross-link compensator. For the technical implementation of the cross-link compensator, a method is proposed for approximating its transfer functions, which makes it possible to significantly reduce their order. All conclusions are confirmed by the results of mathematical modeling of output processes in closed and open control systems.

The article describes the study carried out by the authors of the use of parametric geometric models to reduce the complexity of the design of parts that follow the theoretical contour of the bearing surfaces of the aircraft. Also, an algorithm that allows eliminating the existing ad hoc concavity profile is considered.

The research proposes an approach to solving the problems of stability and controllability of aircraft of the aerodynamic scheme "flying wing" in terms of movement on the ground, which consists in combining landing gear systems into an automatic integrated control contour. The analysis of peculiarities of movement on the ground is carried out, disadvantages and limits of application of available control methods are specified. Development of a dynamic aircraft model for virtual testing of control algorithms is described. The conclusion is made about practical advantages that are expected to be obtained because of the introduction of the integrated control contour of landing gear systems. Are additional possibilities of the proposed approach: in addition to using on unmanned aerial vehicles to operate in automatic control mode, the development can be used on manned and remotely piloted aircraft as an auxiliary system that helps the pilot to withstand the take-off and run direction under random energetic disturbing influences, similar to that used on modern cars of assisting systems such as ESP (Electronic Stability Program) -electronic dynamic stabilization system.

The effectiveness of methods for forming a comprehensive information facility to make targeted decisions in high-tech industries is determined by the systematic organization of processes aimed at measuring costs and results in automated systems. The formation of such information facilities in terms of ensuring unity and comparability of information is considered. The main indicators are the description and language implementation of models of automated quality assurance systems, ways to ensure the unity of information and consistency of requirements, metrics of integrity and comparability of data in systems.

The article is devoted to finding the optimal schemes of fiber placement at the production of constructions, reinforced with continuous fibers by 3D printing method. As the optimization of the objective function one of the criteria for the destruction of the composite was chosen. For the process acceleration of multiple solution of the system of partial differential equations describing the stress-strain state of the structure, a computational algorithm based on wavelets built through subdivision schemes is proposed. To set the local coordinate system, it is proposed to use analytical functions, which will be constructed using the well-known Dini and Cisotti formulas, just by specifying the direction of laying the fiber at the product boundary. The article also presents a lifting scheme (lifting scheme) allowing to construct biorthogonal wavelet systems with specified properties using some initial biorthogonal wavelet systems with filters.

The algorithm of formation to structure of self-regulating technical system, optimum on Pareto, which quality is defined by the plan of measurements from agents, on the basis of kvalikativny functions is considered.