D. Popovic

Thin-layer modelling of black tea drying process

Abstract An experimental dryer was developed for determining the kinetics of black tea drying. Drying characteristics of tea were examined using heated ambient air for the temperature range 80–120°C and air flow velocity range 0.25–0.65 m/s. The data of sample weight, dry- and wet-bulb temperatures and air velocity of the drying air were recorded continuously during each test. The drying data were then fitted to the different semi-theoretical models such as Lewis, Page, modified Page, two-term and Henderson and Pabis models, based on the ratios of the difference between the initial and final moisture contents and the equilibrium moisture content. The Lewis model gave better predictions than other models, and satisfactorily described the thin-layer drying characteristics of black tea particles. The effective diffusivity varied from 1.14×10 −11 to 2.98×10 −11 m 2 /s over the temperature range. The temperature dependence of the diffusivity coefficient was described by the Arrhenius-type relationship. The activation energy for moisture diffusion was found to be 406.02 kJ/mol. Temperature and air velocity dependence on drying constant was described by the Arrhenius-type and Power-type relationships. The coefficients of determination were above 0.996 for both relationships. The Arrhenius-type model was used to predict the acceptable moisture ratios at the experimental drying conditions and to understand better the influence of drying variables on drying rate constant. The results illustrate that in spite of high initial moisture content, the drying of tea particles takes place only in the falling rate period. This single-layer drying equation can be used for the simulation of deep-bed drying of black tea.

Analysis and control of industrial processes

1. Survey Papers.- Review and Future of Adaptive Control Systems.- Expert Systems and Their Applications: A Survey.- Fibre Optic Communication Systems in Industrial Automation.- Route Optimization Algorithms for Expert Systems.- 2. Dynamic Systems.- Enclosure Methods and Their Applications in Control Theory.- Improved Computation of Balancing Transformations for Model Reduction of Minimal Systems.- Adaptive Control of Linear Plants With Unknown High-Frequency Gain.- Design of a Full-Order Observer and its Minimal-Order Version for a One-Link Flexible Robot Arm.- 3. Model Building and System Simulation.- Modelling and Simulation of Non-Isothermal Processes for the Reduction, Precipitation and Characterization of Metal Phases in Zeolites.- Dynamics of Heterogeneously Catalyzed Oxydation Rections on Palladium Supported Catalysts.- Modelling and Simulation of the Methanation from Carbon Monoxide-Rich Synthesis Gas.- On Parameter Estimation of Ecosystem Models.- 4. Process Analysis and Control.- A Petri-Net-Based Tool for Computer-Aided Model Building, Simulation, and Analysis of Engineering Systems.- Control of Distributed Parameter Systems Using Pointwise Location of Sensors and Actuators.- On Modelling and Control of a Reactor for Pyrolysis.- Computer Coupled Laboratory Reactor for the Study of Temperature Profile Dynamics.- 5. Expert Systems.- Some Aspects of an Architecture for Real-Time Expert Systems in Industrial Plant Automation and Supervision.- Rule-Based Modelling of Dynamical Systems.- Multi-Sensory Signal Fusion.- 6. Intelligent Robots.- Goal Oriented Behaviour of Robots for Space Applications.- Basic Principles of Intelligent Task Planning for Autonomous Robot Systems.- Collision Free Motion Planning for Robot-Manipulators.

MODELING OF DESORPTION ISOTHERM OF BLACK TEA

In this paper the relationship between the moisture content of black tea and the relative humidity at different temperature conditions is interrogated. The equivalent moisture contents was measured using the static gravimetric method under the relative humidity range of 10% to 90% and under the temperature range of 25°C to 80°C. The obtained data was analysed using statistical data processing techniques and the results fitted to the different sorption isotherm equations like GAB, Oswin, Henderson, and Halsey at various temperatures. The G.A.B. and Oswin equations show good agreements with the obtained results. The results indicate that the desorption isotherms of black tea are sensitive to the temperature. At constant water activity, an increase in temperature decreases the equivalent moisture content of black tea.

DESORPTION ISOTHERM MODELLING OF BLACK TEA USING ARTIFICIAL NEURAL NETWORKS

ABSTRACT Neural network approach was selected for the modelling of desorption isotherms of black tea and a neural network model trained using Levenberg-Marquardt algorithm. The trained network was used for studying the influence of randomised training data of temperature in the range of 25–80°C and the water activity range of 0.10–0.90 and different number of hidden neurons. The obtained results, compared with the results already obtained with the well known phenomenological GAB model, already tested by the authors [9], have proven to be better.

Backpropagation based training algorithm for Takagi-Sugeno type MIMO neuro-fuzzy network to forecast electrical load time series

Describes a backpropagation based algorithm that can be used to train the Takagi-Sugeno (TS) type multi-input multi-output (MIMO) neuro-fuzzy network efficiently. The training algorithm is efficient in the sense that it can bring the performance index of the network, such as the sum squared error (SSE), down to the desired error goal much faster than that the simple backpropagation algorithm (BPA). Finally, the above training algorithm is tested on neuro-fuzzy modeling and forecasting application of electrical load time series.

Nonlinear combination of forecasts using artificial neural network, fuzzy logic and neuro-fuzzy approaches

In the actual practice, it becomes interesting from the efficiency point of view to combine various forecasts of a specific time series into a single forecast and to interrogate the resulting forecasting accuracy. The combination is usually nonlinear. Various intelligent combination techniques have been suggested for this purpose, based on different neural network architectures, including the feedforward neural network and evolutionary neural network. In this paper, the nonlinear combination of time series forecasts is proposed, based on isolated use of neural networks, fuzzy logic and neuro-fuzzy systems. On some practical examples it is demonstrated that the nonlinear combination of a group of forecasts based on intelligent approach is capable of producing a single better forecast than any individual forecasts involved in the combination.

Mechatronics in engineering design and product development

System technology: sensors and actuators in mechatronics microsensors and microactuators microcomputer technology intelligent controllers communication systems. Design approaches: conceptual design computer-aided design of automotive control systems using MATLAB, Simulink and Stateflow rapid protoyping of mechanical and electronic subsystems of mechatronic products. Design-related issues: system integration optimality of system performance system software. Application-related issues: mechatronic system applications an operators model for control and optimization of mechatronic processes ethics in product design.

Real-time identification and control of a continuous stirred tank reactor with neural network

A continuous stirred tank reactor (CSTR), modelled by two nonlinear ordinary equations with two manipulable inputs, is online identified and controlled using neural networks. An enhanced learning algorithm is used for training the neural network. The multivariable nonlinear controller is derived from a Lyapunov function, which can ensure that the whole system be globally asymptotically stable at both the stable stationary points and the unstable stationary points. An inverse model of the CSTR, which is online built with neural network, serves as the predictor of the control variables. The temperature and concentration in CSTR can be maintained stable around the setting values in spite of the disturbances of the inlet concentration and inlet temperature of the reactants.<<ETX>>

Retaining diversity of search point distribution through a breeder genetic algorithm for neural network learning

Genetic algorithms (GA) have been used for training of fixed structure neural networks and for optimisation of network structure. The crucial issue of algorithms is their premature convergence that deteriorates the diversity of individual search points. Several techniques have being applied to retain the diversity of the search point distribution. In this paper the application of a breeder genetic algorithm (BGA) for neural network learning is considered as well as the problem of retaining diversity. Truncation selection, extended intermediate recombination, and variable mutation range are proposed. It is shown that the performance of BGA is superior to GA in retaining diversity.

Forecasting chaotic time series using neuro-fuzzy approach

A neuro-fuzzy approach for forecasting of chaotic time series is proposed, based on neuro-implementation of a fuzzy logic system with the Gaussian membership functions. To construct the neuro-fuzzy system that will approximate and forecast the future values of a chaotic time series, the parameters of the membership functions, i.e. the mean (c) and the variance (/spl sigma/) of the selected Gaussian functions, as well as the center of fuzzy region (y/sup l/) are to be adjusted either by backpropagation or the Levenberg-Marquardt training algorithm. To examine the effectiveness of the forecasting method the performance function, like the sum squared errors, mean squared errors, and mean absolute errors, are evaluated. In this way it was shown that the proposed neuro-fuzzy approach is an excellent tool for chaotic time series prediction.