Search

Methodology of shape optimization of the frequency converter blades, which are heated by Joule losses due to nonuniform current flow, is presented. For processing this coupled and nonlinear problem, its full mathematical model is replaced by a suitable equivalent model, on which the optimization is subsequently performed. The paper demonstrates the possibilities of acceleration of complete simulations, where it is often problematic to carry out the optimization process, using a suitable equivalent model, which would be able to provide sufficiently accurate results in the initial design stages.

The Mixture of Experts (MoE)–based approach is applied to verify the possibility of using surrogate models for searching the optima of complex multicriteria problems with constraints. This approach can successfully solve problems when the design space is limited by a higher number of constraints and traditional methods of Design of Experiments (DoE) in conjunction with one surrogate model are not able to partition the design space acceptably enough for further prediction. The methodology is tested on a well-known DC motor benchmark, where the electromagnetic and temperature fields were solved analytically, in a simplified form. The Mixture of Experts (MoE)–based approach is applied to verify the possibility of using surrogate models for searching the optima of complex multicriteria problems with constraints. This approach can successfully solve problems when the design space is limited by a higher number of constraints and traditional methods of Design of Experiments (DoE) in conjunction with one surrogate model are not able to partition the design space acceptably enough for further prediction. The methodology is tested on a well-known DC motor benchmark, where the electromagnetic and temperature fields were solved analytically, in a simplified form.

An approach based on recurrent neural networks (RNNs) is applied to verify the possibility of using surrogate models for the prediction of dynamic nonlinear problems. Modeling complex time dependencies is currently still a challenge, when the structure of the neural network needs to be adapted to the dynamics of the problem. In this paper, the possibility of using prediction in space-time problems is illustrated by the possibility of using it to predict the course of the current in a simple RL circuit that is powered by a voltage source.

Control systems for every simulator have to be fast and precise. These are exact properties that are needed when designing a paraglider simulator. With it’s complex set of controls and complicated aerodynamics physics, paraglider simulator requires a very precise system to mimic it’s real world counterpart behaviour. From construction to software solution, a team of people with various sets of skills was put together to develop the control system, that is worthy of simulator going into production. This paper will go over how the control system was developed, what parts we’re chosen for it and what can we expect from such system.

This paper deals with kinematics and the design and implementation of the paraglider simulator drives. Part of this article contains the necessary calculations, that are important for the mechanical design and manufacture of drives. There are given reasons which led to using 3D printing. The project aims to present the design of drives control, which includes the solution to the kinematics problem of the simulator. Finally, the equations determining the length of ropes for a given point of the seat in space are shown.

Nowadays, there is a huge interest in the design of medium frequency transformers. Due to their high power density and relatively good efficiency. These transformers are still used in wind farms and traction, where not only the efficiency, but the overall size and weight of the transformer are important design criteria. However, the determination of the load losses is based on analytical or pseudo-empirical formulations during the design optimization process. In this paper, an open-source robust design optimization framework process based medium frequency transformer design process is introduced. During the solution of this process, the magnetic field is calculated directly by the builtin FEM solver.

The main goal of the Artap project is to provide an extensive infrastructure for robust design optimization, where usually many different numerical solvers have to be used together and the impact of the manufacturing uncertainties have to be minimized. Artap is an open-source software platform, developed jointly with the coupled numerical field solver, Agros Suite. Artap ensures interfaces for a broad collection of optimization algorithms (genetic and evolutionary algorithms, various interfaces to libraries such as Nlopt, Bayesopt, etc .), tools for machine learning (neural networks, Gaussian processes, etc. ), finite element solvers (Agros Suite, Comsol, Multiphysics, deal.II). The implemented tools offers an easy and straightforward solution not only for robust design optimization but parameter identification, model order reduction, and shape optimization, as well. Moreover, Artap provides automatic parallelization of the optimization process. The paper presents the structure of the framework and technologies powering the project. The main features of Artap are demonstrated on an induction brazing process design tasks., Comment: 6 pages, 10 figures, the paper presented in ICDS 2019 and accepted for publication in IEEE Xplore

A linear peristaltic pump is designed and modeled. The pump is controlled by a set of five electromagnetic actuators with ferromagnetic cores pressing in the given order an elastic tube with pumped liquid. The paper shows and describes the basic arrangement of the device, presents its mathematical model together with the numerical solution and compares selected results of simulation with experimental data measured on its prototype.

Methodology of design of a brushless direct-current (BLDC) motor is proposed, taking into account not only common technical parameters, but also robustness, overload, design simplicity and overall costs. Also, basic control algorithms are discussed and tested.

Mathematical model of thermoelastic microactuator with embedded intelligence is proposed, together with the possibilities of its control. The microactuator consists of a metal body with arm. The system is heated by direct current passing trough it, which leads to deformation of the arm by stresses of thermoelastic origin. This deformation is controlled by the source current. As the measurement of the deformation is technologically unfeasible, it is carried out indirectly from the actual temperature of the heated body, using a thermal sensor that controls the input voltage.

A novel way of optimizing a reluctance motor with rotor manufactured by 3D printing technology is presented. The optimization algorithms must take into account the tolerances of its production that are generally higher compared with classical processing, which requires a high level of robustness. The paper presents complete 3D mathematical models of the forward and backward tasks together with the ways of their numerical solutions. The methodology is illustrated with a practical example.

Cogeneration power plants have already been operated in the Czech Republic for several decades. These cogeneration power plants have been mostly operated with original technologies. However, these original technologies have to be continuously innovated during the entire operation time. This paper is focused on one of the possible innovations, which could lead to better sustainability and improved flexibility of the cogeneration power plants. Backpressure turbines are still used in many cogeneration power plants. However, backpressure turbines are currently losing suitability for cogeneration power plants, because they always need sufficient heat demand for optimal operation. Backpressure turbines rapidly lose efficiency when facing a lack of heat demand, i.e., mostly in summer season. Currently, condensing turbines are a preferable option for cogeneration power plants, which generally achieve less e ective operation, as condensing turbines are able to operate with optional heat demand. Therefore, backpressure turbines are often replaced by condensing turbines with regulated outputs. In spite of the current trend, this article will present an innovative topology, which retains the original backpressure turbine with the addition of the organic Rankine cycle for residual energy utilization.

A method of finding the temperature dependences of physical parameters of metals is proposed and discussed. The method is based on the combination of measurement of the surface temperature of the material and subsequent solution of the inverse problem using the multi-parametric optimization. Particular steps and algorithms are illustrated with an example.

The ubiquitous internetworking of devices in all areas of life is boosted by various trends for instance the Internet of Things. Promising technologies that can be used for such future environments come from Wireless Sensor Networks. It ensures connectivity between distributed, tiny and simple sensor nodes as well as sensor nodes and base stations in order to monitor physical or environmental conditions such as vibrations, temperature or motion. Security plays an increasingly important role in the coming decades in which attacking strategies are becoming more and more sophisticated. Contemporary cryptographic mechanisms face a great threat from quantum computers in the near future and together with Intrusion Detection Systems are hardly applicable on sensors due to strict resource constraints. Thus, in this work a future-proof lightweight and resource-aware security concept for sensor networks with a processing stage permeated filtering mechanism is proposed. A special focus in the concepts evaluation lies on the novel Magic Number filter to mitigate a special kind of Denial-of-Service attack performed on CC1350 LaunchPad ARM Cortex-M3 microcontroller boards

The paper is divided into two parts. In the first section the paper deals with shape optimization of high-power converter busbar. The second section is focused on the topology optimization which was applied on the busbar. The main goal of the optimization is to ensure that the power is equally distributed among particular active components.

Numerical modeling of acoustic field produced by a set of common commercial loudspeakers is carried out. The goal is to obtain a time-dependent level of acoustic pressure in the microphone. The continuous mathematical model is derived from the physical laws and is considered for the ideal fluids with properties of a perfect continuum. Results of the model derived from pressure acoustics and data obtained from measurement are compared.

The paper deals with calculation of iso-lines of radiation-field from isotropic rectangular source using pseudo arc-length continuation. The proposed approach is demonstrated on fire-safety application.

Optimization of an induction pump is carried out. Based on numerical simulation, weight-and-size parameters are reduced and thrust of pump is increased. The problem of optimization is realized using the open-source package of numerical modeling Agros Suite. The task is formulated in an electromagnetic approximation, neglecting a non-uniform distribution of a velocity in a pipe. The saturation of the magnetic circuit and back iron is taken into account in a quasi-static formulation. The problem is solved in two-dimensional formulation, taking into account the transverse edge effect by coefficient of Bolton. The effectiveness of optimization algorithms implemented in Agros Suite (BOBYQA, BayestOpt, Limbo, NSGA-2) and Comsol Multiphysics (Nelder-Mead, BOBYQA, Coordinate Search) is compared.

The design optimization of the laboratory induction pump is carried out. The forward task is formulated as a nonlinear magnetic problem that is solved by the professional code COMSOL Multiphysics. The backward (optimal design) problem is solved by Coordinate search, BOBYQA, Monte Carlo, and Nelder-Mead algorithms implemented in COMSOL and also by some other algorithms built in OPTILAB, which is a module of open source code Agros Suite. The methodology is illustrated with an example. The result of optimization is the reduction of mass of selected parts of the pump, while the resultant force acting on the secondary element increases. Benefits of all optimization methods are discussed.

Numerical modeling of harmonic acoustic field produced by a common commercial loudspeaker is carried out The goal is to obtain the directional characteristic of the device. The continuous mathematical model is derived from the physical laws and is considered for the ideal nuids with properties of a perfect continuum. The model of the loudspeaker in the axial symmetry (2D) is described by the Helmholtz differential equation for harmonic acoustic field and for the correct solution the appropriate boundary conditions and material properties are also defined. The data obtained during the measurement must be used to implement the model. In our case, there was the fundamental knowledge of the velocity of the diaphragm oscillation for the selected frequencies.

A novel way of calibration (estimation of unknown material parameters and boundary conditions) of physical models is presented, based on measured data and advanced optimization and penalization techniques. The methodology is illustrated with an example whose results are discussed.

Optimization of the process of induction clamping is presented, consisting in finding the optimum shape of the field coil. The objective function takes into account the required expansion of the bore of the clamping head. While the forward problem is, as a coupled task with nonlinear and nonstationary interaction of three physical fields (magnetic field, temperature field and field of thermoelastic displacements), solved by Finite Element Method, the backward problem is solved using several heuristic and deterministic algorithms. The methodology is illustrated with a typical example.

Electrostatic separation of triboelectrically charged plastic particles is investigated and verified on an experimental device. After presenting the fundamentals of the process in the introductory part, attention is paid to the stucture of the device and functions of its particular elements. The last part of the paper discusses the results of experiments aimed at separation of a mixture consisting of acrylonitrile butadiene styrene (ABS) and thermoplastic copolyester (TPC). Measured are various parameters of the process such as deflections of individual kinds of particles in electric field, efficiency of separation and purity of particular levels to be separated.

The utilization of the cross-correlation technique in pulsed eddy current testing is described and illustrated with an example of an artificial defect in an electrically conductive steel plate. The investigated eddy current probe is of a differential type, consisting of one excitation and two sensing coils with a ferrite magnetic circuit. The numerical solution is carried out in the COMSOL Multiphysics using Magnetic field and Electrical circuit modules. The results are verified by performing laboratory experiment. Important aspects of the cross-correlation in the pulsed eddy current testing are then analyzed and discussed.