Search

The article focuses on the design, construction and manufacture of an inspection vehicle intended to access difficult-to-reach places. The vehicle is able to monitor piping system failures at the view angle of 180° in deep depths and adverse environments. The individual components of the inspection vehicle, more detailed their programming and production on CNC machines are discussed. The vehicle is supposed to easily run into the piping systems and can be safely pulled out. A control system was created for motion and video signal transmission to the operator from above the ground. Aluminum alloy (EN AW 2024) is the predominant material of the manufactured components and at the same time has ideal processing and operating properties. Proposed inspection vehicle is a robust and functional solution with minimal maintenance. © 2018 Trans Tech Publications, Switzerland.

The aim of this paper is to evaluate the usability of the self-organizing migrating algorithm (SOMA) in a nonlinear system predictive control area. The model predictive control is based on an objective function minimization. Two approaches to model predictive control applied on a nonlinear system are studied here. Firstly, the SOMA was used to minimize the objective function, secondly, the fmicon function included in the MATLAB optimization toolbox was used for the same. The nonlinear system simulated here is an exothermic semi-batch reactor mathematical model based on a real chemical exothermic process. Also the input data used here to simulate the process were obtained from the same real process. Results obtained by the simulation means were than evaluated using suitable criterion which was defined for that purpose and discussed. © 2018 The Authors, published by EDP Sciences.

Robotic systems are used wherever human limits are achieved. The inspection robots are used for securing the access to the manholes, pipes and sewage systems, for monitoring faults and for the eventual minor repairs. The paper presents the design, construction and assembly of a four-wheeled inspection robotic rover used in sewage pipes with diameters larger than 200 mm. We managed to design the robust, waterproof and dirt-resistant washable rover solution, which has been proved to be very functional and easy to produce. © The Authors, published by EDP Sciences, 2017.

Increasing urge to raise production rate and production quality in the industry brings new requests and challenges. One of them is demand for accuracy and precision of produced parts. Especially in the CNC machining, where the expectations are high, the companies face new issues. Therefore, it is very important to recognize, understand and cope with the technological factors influencing the production accuracy and surface quality of CNC machined parts.

High speed grinding methods are continuously being developed and have been highly accepted due to high productivity and resulting surface quality when machining materials dissimilar in physical and mechanical properties. Grinding wheels are the most important part of the entire technological system, in particular their stress state. Hence, the article focuses on the determination of the mathematical model of the grinding wheel with uniform strength. Based on it are calculated critical wheel speeds for various bonding material. Then, the optimal models of grinding wheel under high wheel speed are solved by finite element analysis.

The article is focused on the production of permanent casting patterns from wood for impermanent moulds by the 5-axis machining centre. Programming of casting pattern was dealt with CAM software by two different manufacturers. Casting pattern, adapted by appropriate allowances, was loaded in the SolidCAM R10 and the NX 7.5 by Siemens. Ergonomics in the working environment of programmes, operation verifications and theoretical machine time were compared. The casting pattern was machined after generating the NC code on FC3800 MACH milling machine. Monitored parameters during computer programming were associated with the measured values when machining casting pattern. Estimated parameters between the computer simulations differed minimally in general, although the programming methods of cutting operations were different. Considerable differences, as expected, were observed in the values of estimated and actual machine time.

The paper presents a control mechanism design for a semi-batch chemical reactor. The data obtained by chemical engineering analysis of real experiments are used to simulate the semi-batch process. A mathematical model based on the real reactor geometry and size is used to simulate the whole process. The process simulations are implemented in MATLAB / Simulink environment and suitable PID and Model Predictive Control are also proposed. Because of that the chemical reactor is a complex and nonlinear system, the PID controller has to use an online identification to be able to deal with nonlinearities. Results obtained by simulations are compared and discussed. © (2013) Trans Tech Publications, Switzerland.

This paper presents numerical models that are able simulate a buckling of the rubber boot. Rubber boot must be able of a large change of its length. A buckling of cylindrical shape of the boot during this deformation would cause a serious problem. Due to the space limitation in an assembly it is quite difficult to find optimal shape of the boot profile which will be able of the required deformation without the buckling. We have created numerical models that are able to simulate mechanical behavior of the compressed boot including the buckling of the boot with an inappropriate profile. Due to the material of the boot (rubber), a nonlinear hyperelastic material model was used in the analyses. The material constants of this hyperelastic model were obtained from the uniaxial and equibiaxial tests of the boot material. Created numerical models were used to design the optimal shape of the boot which eliminates the risk of the buckling.

The contribution studies prediction of the given semibatch reactor using multilayer feed-forward neural networks. The two prediction approaches are tested-signal prediction approach and system prediction methodology. The first approach is commonly applied in time series prediction, while the input-output models in the second methodology are used for example in the control tasks. Furthermore, the resulting predictor is used for the model predictive control of the reactor in order to test performance of the developed method.

The paper studies time series prediction using artificial neural networks. The special attention is paid to the influence of size of the input vector length. Furthermore, the prediction of standard single-dimensional data signal and the prediction of multi-dimensional data signal are compared. The tested artificial networks are as follows: multilayer feed-forward neural network, recurrent Elman neural network, adaptive linear network and radial basis function neural network.

The paper presents application of Siemens RobotExpert software of industrial robot offline programming. The deburring process of aluminium wheel is described and developed. The robotic work-cell contains robot ABB IRB 1600id and two axes positioner ABB IRBP A 750 D 1000 H 700. The final robot tool path is checked using the collision viewer, the joint status monitor, the tool centre point speed viewer and tracker. © 2013 Published by Manufacturing Technology.

Modeling of complex multidimensional systems is generally difficult especially in cases when a technological process is regarded. What is more, data obtained from technological systems are usually influenced by noise, disturbances and measurement errors. This paper presents case study of modeling of technological process, namely polymer materials grinding, with artificial neural networks.

Hard surface layers of polymer materials, especially polypropylene, can be formed by chemical or physical process. One of the physical methods modifying the surface layer is radiation cross-linking. Radiation doses used were 0, 30, 45, 60 and 90 kGy for unfilled polypropylene with the 5 % cross-linking agent (triallyl isocyanurate). Individual radiation doses caused structural and micromechanical changes which have a significant effect on the final properties of the polypropylene tested. Small radiation doses cause changes in the surface layer which make the values of some material parameters rise. The improvement of micromechanical properties was measured by an instrumented microhardness test. X-ray diffraction was used to study the influence of the structure.

This article introduces an application of artificial neural network with radial basis function in modeling of polymer materials grinding. This real technological process has many input parameters that influence results of grinding. In this paper the two key parameters were selected - feed rate and depth of cut. The task of the artificial neural network based predictor is to provide resulting surface roughness (Ra and Rz).

Artificial neural networks are commonly used for prediction of various time series, linear and nonlinear systems. Nevertheless, the choice of proper type of artificial neural networks is difficult task, because each class of artificial neural networks has different features and abilities. Aim of this paper is to compare and benchmark four typical categories of artificial neural networks in artificial time series prediction and provide suggestions for this kind of applications.

This paper presents usage of artificial neural networks for modelling of laser micro-machining process. Results of the laser micro-machining – surface quality of product and his utility in specific application – depend on the laser-machine parameters and the polymer material type. Commercial CO2 laser Mercury L-30 by LaserPro, USA was used for cutting specimens. This laser system has two parameters - power and feed. The article also shows optimization of the laser micro-machining using artificial neural network. In order to interpret complicated dependencies between technological characteristics of laser micro-machining and output parameters software Matlab 6.5 with Neural Network Toolbox was used. The experimental results were evaluated and depicted into the graphs.

This paper deals with investigation of physical and mechanical characteristics of grinding materials and evaluation shaping of these properties in abrasive tools. Experimental results of high power grinding are introduced. State of stress enumeration of rotating wheels and critical velocity of shaped grinding wheels are shown. Evaluated surface roughness parameters are based on changing technological conditions.Within the cutting process, cutting forces are measured on strain gauge dynamometer. Results are given in comparison with standard metal material C45. Behaviour of plastic materials during the grinding was not so far been experimentally investigated.

The contribution is aimed to the design of a robot hand. The taks of the robot is to load and unload steel parts into CNC lathe machine EMCO TURN 323-II. Three different designs are presented, advantages and disadvantages are evaluated. The best designed is proposed in the conclusion.

The work is aimed to research of predicting abilities of artificial neural networks. The characteristic samples of artificial neural network types were selected to be compared in numerous simulations, while influences of key parameters are studied. The tested artificial networks are as follows: multilayered feed-forward neural network, recurrent Elman neural network, adaptive linear network and radial basis function neural network.

Je ukázáno, že transientní teplotní napjatost vrstevnatých stěn není výhradně způsobována samotným rozdílem v hodnotách koeficientů teplotní roztažnosti jednotlivých vrstev. Vznik transientní teplotní napjatosti je podmíněn jak danou strukturou vrstevnaté stěny a danými okrajovými podmínkami, tak existencí gradientu teploty ve směru normály k ploše stěny. Praktická aplikace je zaměřena do problematiky recyklace PCB desek, kde je snahou dosáhnout vlivem teplotního pnutí separace vrstev. Úloha modelování teplotní napjatosti v této oblasti spočívá v predikci možností separace v procesu recyklace, v závislosti na způsobu teplotního namáhání a materiálových parametrech vrstev., Results of FEM modelling of transient thermal stress analysis in layered walls are given in the article. It is shown that thermal stress alone is not solely caused by differences in coefficients of thermal expansion of individual layers. The emergence of transient thermal stress is subject to both the layered structure of the wall and given boundary conditions, as well as the existence of a temperature gradient in the direction normal to the surface of the wall. A practical application focuses on the issue of recycling of PCB boards with the effort to achieve separation of layers due to thermal stress. Role modelling of thermal stress in this area lies in predicting the possibility of separation, depending on the type of thermal stress and material parameters.

The paper presents different control methods applied on a chemical exothermic semi batch reactor. Because of a strongly exothermic chemical reaction the in-reactor temperature is rising very fast depending on the reaction component dosing. Thus, the temperature control is necessary. The system control is generally difficult because of its nonlinear behavior. To simulate the real process a mathematical model including reaction kinetics was used. Three control methods were simulated - two step control with penalization, PID control and model predictive control. Keywords: Predictive control, PID control, two step control, chemical semi-batch reactor, process modeling.

The paper deals with laser micro-machining of polymer material. The experimental data are processed and modelled by artificial neural network (multilayered feedforward neural network). Furthermore, the new set of experiments was performed to test and verify the obtained neural model. The acquired results are graphically displayed and evaluated.

The goal of this paper is to present interesting way how to model and predict nonlinear systems using recurrent neural network. This type of artificial neural networks is underestimated and marginalized. Nevertheless, it offers superior modelling features at reasonable computational costs. This contribution is focused on Elman Neural Network, two-layered recurrent neural network. The abilities of this network are presented in the nonlinear system control. The task of the controller is to control the liquid level in the second of two interconnected cylindrical tanks. The mathematical model of the real-time system was derived in order to test predictor and consequently the controller in Matlab/Simulink simulations.

Příspěvek pojednává o testování prediktivních schopností umělých neuronových sítí. Zvláštní pozornost je věnována výpočetní náročnosti různých typů umělých neuronových sítí and jejich použitelnost v modelovém prediktivním řízení., This paper deals with the testing of prediction abilities of artificial neural networks. Special attention is paid to computational demands of various types of artificial neural networks and their applicability in model predictive control. In order to provide generally comparable results, the CATS benchmark (Competition on Artificial Time Series) was chosen as the testing method.

Příspěvek prezentuje možnosti aplikace neuronových sítí pro modelování procesu laserového mikroobrábění polymerních materiálů, kdy výsledky procesu jsou závislé především na technologických parametrech použitého zařízení a typu polymerního materiálu. Pro přípravu vzorků byl použit komerční CO2 laser Mercury L-30 , jeho ž laserový systém umožňuje měnit paramerty výkonu a posuvu. Neuronové sítě byly použity pro vyhodnocení komplikovaných závislostí mezi vstupními a výstupními charakteristikami celého procesu, This paper presents usage of artificial neural networks for modelling of laser micro-machining process. Results of the laser micro-machining ? surface quality of product and his utility in specific application ? depend on the laser-machine parameters and the polymer material type. Commercial CO2 laser Mercury L-30 by LaserPro, USA was used for cutting specimens. This laser system has two parameters - power and feed. The article also shows optimization of the laser micro-machining using artificial neural network. In order to interpret complicated dependencies between technological characteristics of laser micro-machining and output parameters software Matlab 6.5 with Neural Network Toolbox was used. The experimental results were evaluated and depicted into the graphs.

In this contribution the three various artificial neural networks are tested on CATS prediction benchmark. The results are compared and evaluated. Furthermore, these artificial neural networks are tested in model predictive control on the t-variant system. The aim of this paper is to present and compare artificial neural networks as interesting way how to model and predict nonlinear systems even with t-variant parameters. The key features of this paper are emphasis of the computational costs of the selected predictors and usage of adaptive linear network which offers short learning times and remarkable prediction error.

This paper deals with model predictive control (MPC) of chemical exothermic semi-batch reactor. A first order chemical reaction is considered to be running in the reactor. The reaction is strongly exothermic so the in-reactor temperature is rising very fast due to reaction component dosing. Thus, the temperature control is necessary. The simulation model of the plant was developed in the MATLAB/Simulink. The system is nonlinear because of chemical reaction kinetics, so its control is difficult by classical methods. The classical MPC objective function was modified in order to improve the control. The WC controller uses an artificial neural network as a predictor.

Generally the artificial neural networks (ANN) are regarded as highly computational demanding method. The usage of ANN in model predictive control as an adaptive predictor is mostly impossible. The aim of this paper is to present and compare one possible way how to reduce computational costs of adaptive predictors based on artificial neural networks. This paper presents real-time system control by two adaptive control methods. The first method is based on the model predictive method with adaptive artificial neural network as a predictor. This artificial neural network offers interesting solution of the computation time problem while using artificial neural network as an adaptive (online) predictor. The second method is established on self-tuning approach. Both these methods are applied to a problem of control liquid level in interconnected tanks. Real-time experiments are performed using Amira DTS200 - Three Tank System. This system is characterized by non-linear behavior.

Cílem příspěvku je prezentovat krátký přehled umělých neuronových sítí, které jsou schopny předikovat chování dynamických systémů. Tento článek se zabývá testováním prediktivních schopností umělých neuronových sítí. Jako testovací signál byl vybrán výstup z nelineárního systému. Tento signál byl zašuměn třemi různými stupni šumu za účelem otestovat vliv šumu na kvalitu predikce. Výsledky jsou srovnány s využitím chybové funkce., The aim of this contribution is to provide short review of artificial neural networks that are able to predict dynamical systems. This paper deals with testing of predictive abilities of artificial neural networks. The nonlinear system output was selected as the testing signal. The signal was noised by three levels of noise in order to test the noise influence to prediction quality. The results are compared using the prediction error function

V modelovém prediktivním řízení s využitím umělých neuronových sítí je zpravidla používán off-line trénovaný prediktor. Avšak tento přístup má nevýhodu v neměnném prediktoru a je nepoužitelný pro t-variantní procesy. Tato nevýhoda může být odstraněna využitím adaptivního prediktoru, který je v případě umělé neuronové sítě naneštěstí neúměrně výpočetně náročný. Navržená metoda - ADALINE významně snižuje výpočetní náročnost modelového prediktivního regulátoru s využitím adaptivního prediktoru na bázi umělé neuronové sítě. Zkoumaná metoda je experimentálně testována a srovnávána s jinou adaptivní řídící metodou - samočinně se nastavujícím regulátorem., Model predictive control using artificial neural network is usually applied with unchanging off-line trained predictor. However, this technique has a drawback in static predictor and it is useless in case of t-invariant processes. This pitfall can be removed by adaptive predictor which is unfortunately, in case of artificial neural network, immoderately computationally demanding. Proposed method - ADALINE significantly reduces computational demands of model predictive control using adaptive predictor based on artificial neural network. Investigated method is tested in experiments and compared to another adaptive control method - self-tuning control.

Cílem tohoto článku je presentovat ADALINE jako adaptivní prediktor, přičemž zvláštní pozornost je věnována použitelnosti v nelineárním modelovém prediktivním řízení. Tato schopnost je demonstrována pomocí dvou simulací v softwaru MATLAB a Simulink. První simulace předvádí predikci měnícího se harmonického signálu. V druhá simulaci je testována predikce výstupní veličiny nelineárního SISO systému s proměnlivými parametry., The goal of this paper is to introduce ADALINE as an adaptive predictor. Special attention is paid for applicability in nonlinear model predictive control. This ability is demonstrated by two simulations in the software MATLAB and Simulink. The first simulation presents prediction of changing harmonic signal. The second simulation describes prediction of nonlinear SISO system output also with parameter changes.

Tento článek se zabývá modelovým prediktivním řízením za použití umělých neuronových sítí. Dále ukazuje možnosti adaptivní lineární sítě při predikci výstupu nelineárního systému a její možnosti pro použití v prediktivním řízení těchto systémů. Jako srovnávací metoda jsou použity vícevrstvé sítě s dopředným šířením., This paper deals with model predictive control using artificial neural networks. This paper shows ability of adaptive linear network to predict nonlinear system output and to be used for predictive control of these systems. Multilayer feed-forward neural networks are used as a comparative method.

Tento článek pojednává o prediktivním řízení pomocí umělých neuronových sítí. Po krátkém úvodu do základů metody prediktivního řízení založeného na modelu, příspěvek prezentuje použití adaptivní lineární umělé neuronové sítě pro prediktivní řízení silně nelineárního systému. Tento článek demonstruje schopnost adaptivní lineární sítě predikovat výstup silně nelineárního systému, což ji činí použitelnou pro prediktivní řízení těchto systémů., This paper deals with predictive control using artificial neural networks. After short introduction to the basics of the model predictive control approach, this contribution presents an application of adaptive linear neural network for predictive control of strongly nonlinear system. This paper shows ability of adaptive linear network to predict nonlinear system output and to be used for predictive control of these systems.