main-content

## Über dieses Buch

The present volume contains the papers which were accepted for presentation at the 3rd International Symposium for Systems Analysis and Simulation held in Berlin (GDR), September 12-:-16, 1988. It is already a tradition to meet a broad international community of experts in systems analysis, modelling and simulation at this symposium. This fact shows the requirements for a forum of presentation and discussion of new developments and applications of modelling and simulation in systems analysis. To realize the great interest in this field one has to take into consideration the developed role of computer simulation as a powerful tool of problem solving. More and more areas in sciences and production have been investigated by mathematical models and computer' simulation. Biological sciences and social sciences are even by now influenced by this trend. The model use on the computer has been very much improved in decision support systems. Parallel simulation will provide drastic shortening of computing time. Parallel simulation and model based decision support systems are brought in the focus of international activities. Numerical mathematics, systems theory and control sciences provide with algorithms supporting the modelling process itself based on simulation or analytic methods. Such simulation systems equipped with tools for modelling and graphics for representing results are real model support systems. A new important impact comes from artificial intelligence by knowledge processing. Expert systems may help decision making in case of missing mathematical models. Expert systems may also support teaching and using simulation systems.

## Inhaltsverzeichnis

### Review of Applied Informatics in Environmental Protection

An effective protection of our environment is highly dependent on precise information on environmental state and development. In this way, computer use in environmental agencies became more and more important. The state of environmental computing in practice is documented in the “Study on DP-Applications in the federal and state Environmental Offices” by the Environmental Federal Office of the F. R. G. (see /18/). It analyses the standard of environmental protection software on the basis of an empirical survey in more than 50 state agencies. The results of the study show on one hand a rapid increase of DP-activities in the environmental agencies during the last years and an intensified use of modern and powerful DP-concepts (such as interactive systems, data base systems, distributed processing with PC-networks or computer colored graphics). On the other hand, the introduction of overall concepts for the development of environmental DP-procedures in the sense of comprehensive, task oriented environmental information systems is still at the beginning. Scientific activities in environmental computing have been rather rare. Most of the environmental computer applications stem from practical projects or other scientific fields without participation of informatics. As a scientific application field for informatics, the environmental domain was only of very limited research interest, although already more than 10 years ago a first conference on this topic was held (see /10/). In the following years only few publications on this subject came out. Just recently an increasing research interest in environmental applications of informatics can be observed as shown by intensified publishing (see e.g. /17/ or /4/) and recent conferences (see /2/, /3/, /7/, /14/, /15/ or /16/).

Bernd Page

### Application of Optimal Control Methods to Simulation of Nature Processes in Arid Zone

In scopes of general system approach to the study of ecology — economical processes of the arid zone we propose the approach to simulation of the spreading processes (desertification, salination of lands etc,) under human impact.

G. A. Dorrer, O. G. Ovezgeldyev, M. G. Dimitriev, M. R. Mansimov

### A Repetitive Working Irrigation Algorithm Using Actual Climate Data

For a long time biologists and scientists in agriculture are engaged to develop growth models (see e.g. van KEULEN, 1986: AUGUSTIN, SCHMIDT, 1980). Descriptive models have the benefit to be as true as possible in detail. Control models represent as correct as possible the input-output behavior and contain the control parameters explicitely. The latter one is suitable for design of control strategies but it is more difficult to build it directly. For this end a deskriptive model can be useful as original substitute. In this paper a control model for sugar beets has been available (HOFFMANN, 1980) and also a soil water model (KOITZSCH, 1977), which enables to ascertain the water supply of the plant (called waterstress) with respect to the water amount which comes from rainfall and/or irrigation. Both models have been developed by the Academy of Agriculture of the GDR.

H. Puta, D. Gerecke

### Simulation of Urban Air Pollutant Processes and Monitoring Stations Location Problem

Investigations into air pollution problem are mainly related to modeling, predicting the pollutant concentrations and monitoring stations location and have been considered to be very important in recent years since the existence of high level concentrations for a long time is injurious to the health of the populous.

Józef Korbicz, I. L. Parmaste, Piotr Gawlowicz

### Structural Analysis of the Climatic Response to a Nuclear War

Global and indirect consequences of nuclear warfare, which would be mediated by the atmosphere, are discussed already since the fifteeth. A definite assertion about the sign and magnitude of potential climatic effects/ however, was not achieved for long. Apart from quantitative investigation, crescent warnings of a possible climate catastrophe after an all-out nuclear exchange originated in the late seventeeth from a general increase in understanding the structure and dynamics of complex systems, from the quickly developing knowledge about paleoclimatic changes, and from global energy balance studies of climate variability and stability. A serious, quantitatively based consideration of atmospheric, climatic and environmental consequences of nuclear war became available in 1982 with an international study [1], which also brought forth the startling discovery of a potential severe atmospheric impact of smoke. Due to propitious coincidence with major attention devoted to Martian dust storms, the “Nuclear Winter” theory could be quickly carved out and further developed as a well-founded scientific reasoning [2]. On the other hand, general indeterminisms (the “scenario”), poorly understood climatic feedbacks, and insufficient self-consistency and resolution of current-generation climate models span a serious range of uncertainty. Activities are worldwide in progress to narrow these margins [3]. One question is often posed in this context: What methods and what general properties of the climate system may the computational analyst rely on in this field?

P. Carl, G. L. Stenchikov

### PEMU — An Impact Model Based Environment Protection Decision Support System

PEMU is being developed as a user friendly decision support system for environmental conservation an a national level. It supplies connections between pollutant emission of point sources, immission (conc./depos.) at receptor points or grid elements and environmental consequences. The user is provided with a tool for evaluation emission reduction strategies by scenario technique and by optimization facilities in order to calculate optimal timing of control strategies and minimal needed immission reduction to meet future goal states of environmental systems. Presently, PEMU is based on S02 and pine stand decline. Inclusions of dust, nitrogenium species, summer/winter immission and spruce stand impacts are being prepared.

A. Sydow, K. Bellmann, R. Straubel, I. Imming, H. Kaschenz, U. Damrath, G. Hofmann, S. Anders

### The PEMU/Air Pollutant Transport Model Based on Emittent-Receptor-Point-Transmission Calculation

PEMU (Prognosis and Decision Support Model for Environmental Protection) has been designed to be especially useful to decision makers involved in SO2 — emission control in the GDR. PEMU deals with a) pollution generation from point sources Csubstances: SO2 in preparation nitrogen, dust) b) pollutant transport processes and c) environmental impacts (forest decline and productivity). The general PEMU — concept is described in [1], an overview about the forest decline model is given in [2] in this volume. In this contribution we describe the implementation and use of part a) and b).

P. Lasch, N. Model, K. Bellmann

### The PEMU Forest-Impact-Model FORSTK A Pine Stand Decline and Wood Supply Model

The simulation model FORSTK is an impact component of the national Decision Support, System for Environmental Protection (PEMU/Air) in order to test consequences of SOP-immission eduction strategies for the future dynamics of pine forest stands with respect to productivity and stability variables by means of scenario and optimization calculations.

K. Bellmann, P. Lasch, G. Hofmann, S. Anders, H. Schulz

### Water and Matter Dynamics in Agro-Ecosystems — a Special Cooperative Project of the German Research Foundation at the Technical University of Braunschweig —

The main objective of this major program is the analysis and forecasting of effects which farming and natural properties of specific sites have on energy fluxes, water fluxes, matter dynamics, microclimate and soil organisms. The differentiation of water and matter dynamics in space and time is analyzed. Special emphasis lies on the investigation of the long-term changes in agro-ecosystems. The atmosphere, the soil surface, the saturated and the unsaturated zones of the soil are the main compartments to be analyzed. Therefore all investigations are focusing on the soil.

J. Richter

### Automatical Control Strategy of Non-Technical Systems in the Groundwater Zone

In Central Europe, in particular in the German Democratic Republic, soil and groundwater zone are intensive and repeatedly used.

P.-W. Gräber

### Modelling and Forecasting of River Flows Using Stochastic Models

The river flow process is a very complex process that involves a number of primary as well as auxiliary variables such as rainfall, groundwater and geomorphologic characteristics of the basin. Because of the complicated nature of the process, it is hard to quantify the contribution of different factors on the flow. In this situation modeling of the river flow process by a stochastic model is recominanded.

Th. D. Popescu, S. Demetriu, V. Marinescu

### Application of Trajectory Decomposition Method of the Water Quality Control Problem

This paper deals with the problem of determining optimal location of Waste discharge points along the whole length of the river and with optimization of waste load quantities discharged into the river. Optimization is made in order to ensure a required quality of river-water.

Mirosław Bereziński, Grażyna Petryczek

### Computer-Aided Decision for Water Quality Management

The complex, system analysis of water quality models is introduced with software support by the system CANDYS. The model is a system of ordinary deterministic or stochastic differential equations with a great number of free parameters. The quantitative validation of the model (parameter fitting), the qualitative validation (validation of stability, limit cycles, bifurcation etc) and statistical analysis of the stochastic model (time dependent state distribution, statistical parameters, significance of statements etc) is done.

P. Braun, P. Rudolph, K.-F. Albrecht

### Pareto-Optimal Cost Division for the Design of Sewage Water Treatment Plants by means of the DSS REH

The achievment of regional water quality goals often involves substantial capital investments and changes in public attitudes concerning resource management. The economic impacts may include not only the costs of facilities designed to reduce the discharge of contaminants into natural surface waters or to improve the quality of waste-receiving waters, but also any limitations on economic development in a particular river basin. Those responsible for the formulation and approval of water quality plans or management policies must have means of estimating and evaluating the temporal and spatial economic and environmental or ecological impacts of these plans and policies. This need has stimulated the development and application of a wide range of mathematical modeling and optimisation techniques for predicting various physical, chemical, biological and economic impacts of alternative pollution control schemes.

A. Gnauck, P. Rathke, R. Straubel, A. Wittmuss

### Two-Level Heuristic Algorithm for Parallel Machines Scheduling

In automated manufacturing systems we often try to minimize work-in-progress, the main part of which consists of in-process inventories necessary for normal functioning of the system.

Jakub Gutenbaum

### The State Variable Diagram of a Hydraulic Control System

This paper suggests a new method of pictorially describing the mathematical model of a hydraulic system by combining the advantages of block diagram and bond graph. The fundamental components are equal flow junction, equal effort junction, power transmission line, control line and various elements. The relationships between the variables are written in the block of every element. The differential relationship is used, so that is convenient to list the state equations from the diagram. A valve controlled motor system is used as an example so as to illustrate the procedures of drawing the state variable diagram.

Shichang Zhou, Xinming Cao, Guiyou Yao

### Decision-Making Simulations with the Micro-CED Programming System

In agreement with the economical and social directives for the years 1986-90, analyzed and approved of in the Third Congress of the Communist Party of Cuba, a considerable number of microcomputers has been installed in different high schools and higher education centers in our country, with the objective of providing access to computer techniques to instructors and students, as well as to improve the efficiency and quality of the educational process.

Victor Giraldo Valdés Pardo

### Application of a Hierarchical Computer System for Picosecond Technology

The unique properties of pico- and especially femtosecond laser light pulses allow for the first time a deep insight into the nature, into the temporal evolution of some of the most fundamental processes in matter, such as the free decay of molecular vibrations, electronic relaxation processes in small and large molecules as well as in solids, phonon decay and energy migration in solids, charge transfer processes and other non-radiative transfer processes, elementary processes of photosynthesis and-of other biological phenomena. Moreover, picosecond technology has provided new possibilities for the manipulation of photophysical and photochemical processes and it is necessary for a number of modern technologies such as the heterogeneous catalysis, making extremly fast electrooptical components, e. g. switches, modulators and receivers, for different kinds of modern electronics, optical computing technique and for biological technology and edicine. This situation makes high demands on the scientific instruments in this field, especially because the majority of the users of these instruments are no “lase-scientists”. On the other hand, a great number of such instruments is the basis for the investigation of phenomena, not yet fully investigated. This report is to show how we in the Centre of Scientific Instruments approached the problem of optimized equipment for various applications with very high flexibility and convenience by means of modern microprocessor electronics.

S. Geschke, N. Kempe, H. Schlawatzky

### Some Comments on Constituents of Contemporary Information Systems Development Methodologies

Development of complex information processing systems in business organizations by interdisciplinary designing team requires specific guidelines — the methodologies, enabling the system developers to transform user’s needs into operating system accomplishing the assumed criteria. In the last two decades numerous information systems development methodologies (ISDMs) has been proposed. Their heterogeneity inclines one to comparisons suggesting the proper selection of tools for practitioners.The limited scope of this paper allows. only to signal certain chosen questions on this subject. After choracterization of the methodological diversity prerequisites, the general Classification of methodologies and specification of their constituents are presented. According to this pattern the introductory specification of solutions relevant to each component is brought forward. The particular emphasis is laid on choice of methods, techniques and computerized tools ibr pragmatic methodology.

Stanisław Wrycza

### Model Based Default Data Generation For Net Systems

When realizing large databases, the main difficulties arise in the collection of data. In general a great part of data is not available on account of man power capacity for administration, due to the difficulties inherent in the measuring methods or even to the lack of time for gathering the data. Therefore, the goal is intimate to reconstruct missing data based on available ones — in other terms —, to reconstruct the missing data on the basis of implicit informations inherent in the data base.

D. Singer

### Various Unitary Based Robustness Studies on a Furnace Control System

In any model-based controller design procedure the engineer is commonly confronted with some significant model uncertainties, consideration of which is relevant to the design result. Frequently, the plant is not completely known, since the only information is based on measured step responses, either a simplified model is used to deal with a complex decentralized control problem, or some plants cannot be described by a linear model at all.

U. Jumar

### Optimisation Problem for By-Pass Routes Creation in the Case of a Teletransmission Line Break-Down

Determination of by-pass routes in teletransmission network with broken trunk-line is investigated. These routes are created from group segments switched in nodes. Technical means used for them are taken away from nonaffected links. Since the grade of service in these links decreases, the right choice of by-pass routes is of prime importance.

Andrzej Winnicki, Grzegorz Płoszajski

### Substitutional Routing for Trunk Line Break-Down in the Teletransmission Network

Trunk-line break-down in teletransmission network interrupts some number of the routes of some different links. It makes necessary to restore, at least partially, failed transmission possibilities for each affected link. Usually it can be done by creating suitable substitutional (by pass) routes from the technical means taken away from neighboring links, not being affected by the break-down.

Kornel Wydro

### Object System Structure and Behaviour Analysis in Database Design

During the last decade, conceptual modelling became — much influenced by the ANSI/SPARC and the ISO TC97/SC5/WG3 proposals — a part of the database, resp.

Martin Holeňa

### Computer-Aided Elaboration of Research Strategies — The Expert System PHARMEX

The search-for strategies-for well directed basic research and applied research is a complex task based on detailed information about different social spheres. Besides, the time factor is a crucial one to the future effect of the cycle Science-Technology-Production-Application.

Karin Schiele, Manfred Koch, Margrit Nachbar

### A Simulation-Based Optimization Algorithm for Operations Control in FMS

This paper is an attempt to model and partially solve the scheduling problem in the environment of modern flexible manufacturing systems. In order to obtain a problem-solving, we propose an efficient algorithm based on simulation.

Carmen Bobeanu, Gabriel Neagu, Florin Filip

### Simulation of Manufacturing Systems Goals, Methods, Tools, and Problems

Simulation models are built arid used in situations where experiments on the real world are not possible or not recommendable. Manufacturing systems are systems of this kind. Experiments are expensive, time consuming and often dangerous. Consequently they have become one of the most often handled and quoted application fields of simulation. The authors, involved in two research projects of Flexible Manufacturing Systems (FMS) Simulation, report about goals, methods, tools, and problems.

Peter Lorenz, Juri Tolujev

### Modelling and Simulation of Automated Welding Manufacturing System

System Analysis and simulation take place in a process acc. to Fig. 1. The starting point is the object to be modelled acc. to a GAD/CAM — problem. It is designated in Fig. 1 as a reality or an original with a given problem. The final result of the process is a usable computer model COMOD which is suitable for carrying out simulation experiments in accordance with the problem to be solved. According to Fig. 1 modeling is regarded as a model development process in clockwise direction and as a model validation process in counter-clockwise direction and computer simulation is a means for assisting the processes /1/, /2/.

W. Krug, F. Blume, F. Erfurth

### On the Application of Advisory Systems and Simulation Processes for the Planning and Control of Manufacturing Systems

An efficient flexible automation of manufacturing systems requires the realization of computer-integrated systems. It is recommended to plan and to control manufacturing by aid of cooperating expert and advisory systems realized at a computer network.

Gerhard Voigt

### The Non-Stationary Distribution of Number of Calls in the M/G/1 Processor-Sharing Queue

The M/G/1 queue with egalitarian processor-sharing is studied by the way proposed in [3]. The new results include the solution of the problem of determining the time-dependent queue-length distribution in the form of Laplace transform of a generating function. We also obtain the joint equilibrium distribution of the conditional sojourn time for a call with a length x and the number of calls in the system at its departure.

S. E. Yashkov

### Robot Manipulator Systems: Analysis and Control

Manipulator systems are rather complex and highly nonlinear which makes difficult their analysis and control. Classic system theory is veil known, however it is inadequate in the presence of strong nonlinear dynamics. Nonlinear controllers produce good results [1] and work has been done e. g. relating the manipulator nonlinear dynamics with frequency response [2–5]. Nevertheless, given the complexity of the problem, systematic methods which permit to draw conclusions about stability, imperfect modelling effects, compensation requirements, etc. are still lacking. In section 2 we start by analysing the variation of the poles and zeros of the descriptive transfer functions of a robot manipulator in order to motivate the development of more robust (and computationally efficient) control algorithms. Based on this analysis a new multirate controller which is an improvement of the well known “computed torque controller” [6] is announced in section 3. Some research in this area was done by Neuman [7,8] showing tbat better robustness is possible if the basic controller structure is modified. The present study stems from those ideas, and attempts to give a systematic treatment, which results in easy to use standard engineering tools. Finally, in section 4 conclusions are presented.

J. A. Tenreiro Machado, J. L. Martins de Carvalho

### Dedicated Computer System for Robot Manipulators

In the last decade, robot dynanic conputational algorithm have been steadily improved. Lagrangian [1] and Newton-Euler [2,3] recursive nunerical algorithns, Horak’s mixed calculation [4], Lisp based symbolic derivation of the robot closed forn equations [5,6] and recently, costumized conputing algorithns [7–9] have been important stages towards nore efficient procedures, Nevertheless, the still existing high computational loads, have prevented the industrial application of such algorithns.

J. A. Tenreiro Machado, J. L. Martins de Carvalho, Antonio M. C. Costa, Jose S. Matos

### Performance Approximation of Robotic Systems Using Rendezvous Networks

One of the challenging topics in robotics is the performance evaluation of robot control software. In complex multisensory robotic systems the control software comprises a large number of tasks executed concurrently on different processors, with real-time constraints [2]. The tasks communicate in synchronous manner in order to prevent unexpected large delays, buffer overflows, etc. [1].

J. W. Miernik, M. B. Zaremba

### Solution of Parameter Optimization and Control Problems in Thermal Systems by means of a Local Adaptive Filter

An attempt is made to develop a single procedure for solving parameter optimization and control problems in thermal systems. An adaptive iterative filter algorithm is considered, suited to parameter identification with the use of special-purpose computers. The problem of controlling the granulation of copper smelting slag in drum-type crystallizers was solved, and the results of the solution are presented.

Yu. M. Matsevity, A. V. Moultanovsky

### Optimization of Heat Delivery to Apartment Buildings Using an Improves Building Model

In Utrecht, the company PEGUS operates some 100 large apartment buildings (about 10000 apartments). The thermal power (primary energy) is provided by the Combined Heat and Power Station (CHP) in Utrecht. There is a deep believe that by lowering the installation water temperature during the night decrease the heating cost can be reduced. In fact this leads to a huge morning peak in power consumption. Because of the higher cost of power during the peak hours the total heating cost can even be higher than keeping the installation temperature constant during the whole night. The aim of our research was to elaborate an optimal control policy avoiding the morning power peak (so reducing energy and costs) meanwhile satisfying some temperature requirements concerning the apartments.

M. Vajta, J. E. Rijnsdorp, S. F. van der Meulen, J. F. C. Verberne

### Performance Measurement and Mathematical Model Identification of the Hydraulic Servo Control System of a Vacuum Electric-Arc Furnace

The static and dynamic performances of the hydraulic servo control system of the electrode in a VHD vacuum electric-arc furnace have been measured. There is a large dry friction area in the performance of cylinder, which is provided by the vacuum seal on the electrode. Through theoretical analysis, the block diagram of the system is obtained. The dry friction can be represented by a negative feedback of cylinder velocity and is proportional to the unit velocity vector of cylinder. For the identification of the mathematical model of the system, a pulse signal is applied into the electrical panel and the responses of the spool displacement of the servo valve and cylinder velocity are recorded under open-loop condition. By discretization of the response curves r the Laplace transforms of valve displacement and cylinder velocity can be calculated. Then the transfer function of the servo valve and the cylinder can be derived by curve fitting. In the procedure of derivation, the dry friction is treated as the unit velocity vector feedback.

Shichang Zhou, Xinming Cao, Chengrui Gao, Yanyong Yang

### Identification of the Mathematical Model of a Servo Valve by means of a Square Wave Signal

Paper describes the experimental results of identification of a two-stage servo valve with moving coil force motor by means of a square wave signal and the method of mathematical treatment. The transfer function of the valve identified is obtained as $$G(s) = \frac{{0.04767}}{{\frac{{{S^2}}}{{{{59.4}^2}}} + \frac{{2x0.47}}{{59.4}}s + 1}}$$ The step response of this transfer function was computed by digital simulation and was compared with the experimental results of the valve. Two lines withing 5% of one another.

Shichang Zhou, Xinming Cao, Chenrui Gao, Yanyong Yang

### Analog Simulation of Transient Behavior of Converters with the Application of Dynamic Thyristor Models

Simulation of transients in converters of electrical energy is useful when designing and examining systems operating at a high frequency, especially those with internal commutation. Transients can be simulated by different methods — dynamic models of semi-conductor valves can be used for this purpose. In analog modelling, parametric models are especially promising. A synthsis of the model can be made by the branch method, adapted to analog technology by J. Deskur [2]. The method, based on topological analysis, permits an easy arrangement of an analog model of any branched electric circuit with diodes and thyristors.

Jan Piłaciński

### Design and Simulation of Advanced Control for a Pulp Digester House

It is reported about the development of advanced control algorithms which are to be implemented in a computer control system for a pulp digester house. The goal of these algorithms is to control the production rate, to improve the quality of pulp and to reduce steam consumption. The control functions considered in the paper include the following: batch sequencing, digester heating control, modification of cooking temperature, prediction of pulp quality during the cooking phase, etc.

N. Hvala, S. Strmčnik, J. Černetič

### Modelling and Simulation of a Drive Motor System

The paper deals with modelling and simulation of a laboratory motorstand. The motor of 1.5 KW, two regulators PI and the observer of the anguIar speed are taken into consideration in the research trial. The computer simulation makes a preliminary investigation to utilize the speed observer in a high power industrial drive system.

Irena Jaworska, Zbigniew Szulc, Grzegorz Kryger

### A Simulation Model for Aluminium Electrolytic Cells Concerning on Process Control

This paper describes the steps and the results of developing work in the field of mathematical modelling and dynamical simulation regarding aluminium electrolytic cells at Hungarian Aluminium Corporation (HUNGALU). Based on the main physical-electrochemical equations, a complex model has been developed. For the examined period of time, the alumina balance formation and the cell voltage alterations are the fundamental processes. The model parameters has been tunned for a real cell. In accordance with the plant situation, the regular operational events as well as the personnel’s interactions can determine the simulated process. The computer-based simulator can be applied for test and educational purposes. The research worker or a trained potroom assistant can watch and modify the forming process. It is an effective method to check the results of possible work routines. With the aid of the simulator, the planned control algorithms can be evaluated. Outlines of theoretical background, with examples of practical applications, are given in the paper.

Laszlo Tikasz, Miklos Zaymus, Andras Cseh

### A Computer System for Decision Support in the Sugar Cane and its Derivatives Industry

In the last years, sugar prices has been decreasing in the world market and its seems this situation will not change. Under this condition, the sugar cane derivatives industry plays an important role in the national economy of producer countries. With aims of analysis of integrated productions of sugar and cane derivatives, the computer package SANPAD is designed: which, based on the Revised Simplex algorithm of linear programming, allows the technical and economical optimization of production strategies. In this paper, the version 1.1 of this package is presented.

Raúl Sabadí Díaz, Eugenio de la Vega, Diego Fernandez

### Simulation Model of High Pressure Gas Pipeline Networks

High pressure simulation of gas pipeline networks is one of the factors which enables to aid the system dispatcher decisions. Because of its complexity, this problem can be solved effectively only by using a computer techniques. As a basic fact one should recognize the dynamic phenomena appearing in the network, accumulating of the gas and a great number of admissible decisions. Moreover in order to elaborate proper control decisions it is necessary to perform a large number of simulations. Each of this simulation is connected with the necessity of the iterative solutions of nonlinear differential equation systems describing the dynamic phenomena appearing in the network. In the papers [4], [5] the solution methods of the above mentioned problem in the case of quasi static flows are described. This algorithms ensure a satisfactory concordance between the results obtained from simulation with the measuring data carried out on a real network especially in case of low pressure gas networks. Only in such situation it is possible to neglect in the model of network the phenomena of gas accumulation.

Ryszard Klempous, Jan Nikodem, Jedrzej Ulasiewicz

### Chosen Problems of Air Traffic Control Systems Modelling

Air traffic control systems /ATC systems/ belong to the class of complex systems with hierarchical structure of control. ATC system consists of two main parts: Air Traffic block /AT block/ representing air traffic carried in the air space and Air Traffic Control block /ATC block/ which includes everything which concerns air traffic control, Air traffic is represented by a set of aircrafts moving in limited air space. Each air-craft is characterized by its specific technical characteristic. The structure of the limited air space is specified by adequate international law regulatins. Air traffic constitutes controlled plant. The control in ATC systems depends on controlling of the current traffic situation and making up appropriate decisions aiming at safety of all air flights carried on in the space being under control.

Andrzej Urbanik

### Design of Multivariable Control Loops for Distillation Columns Using Digital Simulaton

Building a dynamic nonlinear model of a distillation column using submodels for the equipment of chemical technological processes. The submodels are envolved in a library of a digital simulation system for continuous processes. Simulation experiments with the nonlinear dynamic model for to find out the most suitable structure of the multivariable control loops. Derivation of a linearised dynamic model for the use of a program for design linearised multivariable control loops. Realization of the control loops with the TDC 2000 control system.

Ch Hille, M. Wagner

### An Integral Look at Automation Functions in Modern Production Systems

Today’s production systems are continually being upgraded as regards productivity, added-value content and efficiency, with the accent increasingly on cutting costs, saving materials, energy and human effort, and using ecologically sound methods with a minimum of waste. The-extraordinary capital outlays involved require that plants he used to maximum capacity, while greater demands are being made on mobility and flexibility. There is not only a chance but also an urgent need to solve these problems by utilizing microelectronics and informatics to the largest possible extent and optimizing automation strategies.

Herbert Ehrlich

### Dynamic Simulation of Corn Growth

There are several aspirations to model the soil-plant-atmosphere system. They are different efforts according to the order of magnitude of resolution of physical and physiological processes and requirements of computer technique. Our main purpose was to build up a corn model using a personal computer. Because of the importance of maize /Zea mays L./ in Hungary the model was fitted and tested by maize data /c.v. SZTC-255/ derived from the Agrometeorological Research Station at Keszthely /46°44′N; 17°15′E/. The model is based on the Sirotenko model [1], but it is a simpler one.

Z. Dunkel, M. Hunkár

### Studies on the Adaptation of a Winter Wheat Model

The TRITSIM model was tested under Hungarian climatic condition. Winter wheat above ground biomass was simulated using measured and evaluated environmental variables. The results were compared to field experiments.

M. Hunkár, F. Zemankovics

### Modelling of Growth Processes in Broiler Production Using the Simulation System SONCHES

The basic structure of a mathematical growth model for computer aided controlling of the broiler production process is described. The core of this model is represented by the mathematical description of the growth of the single animal under optimal environment. The whole model contains additionally the influence of the animal stock and climatic control. Model building and validation is carried out by means of SONCHES.

K. Grosskopf, E. Matthäus, M. Flechsig

### Computer-Aided Modelling and its Application in Economy and Ecology

Two possible trends in computer aided modelling are considered: 1.Use of the dialogue to consider the a priori information available on the side of user2.Automatic model selection in the way a.of generation of model variants with growing complexity and selection of models of optimal complexity andb.of selection of the best model as a special case.

J. A. Müller

### Policy Simulations and Optimal Control with a Small Econometric Model for Austria

In order to assess the effects of government stabilization policies upon an economy, it is necessary to have some quantitative estimates of the relations between the main macroeconomic aggregates for the country under consideration. Usually this information is provided by an econometric model estimated on empirical data. Moreover, if the question of policy design is addressed, it is desirable to have a standard of comparison for the evaluation of different policies. This can be obtained from simulation and optimization experiments performed with the econometric model? for the latter purpose, an objective function has to be formulated. Policy-makers or their advisers can then be informed about the consequences of particular policy measures and can even be guided such as to choose tnose measures in an optimal way. The present paper intends to provide a first basis for such informations concerning the Austrian economy by estimating a small econometric model for this country and reporting about the results of some policy simulations and an optimal control experiment obtained with this model.

Reinhard Neck

### Total Pollution Effect and Total Energy Cost per Output of Different Products for Polish Industrial System

For many years a broad use has been made of the indices of total energy requirements in the whole large production system corresponding to unit output of particular goods (Boustead I., Hancock G.F., 1979). These indices satisfy well the requirements of system analysis. Because of their efficiency, an attempt was made to solve a similar problem of the determination of the indices of total pollution generation requirements emerging as a result of complex multi-stage (multi-object) production processes.

Henryk W. Balandynowicz

### A Multisectoral Econometric Input-Output-Model of the G.D.R. for Prediction and Dynamic Simulation of the National Income

For analysis, prediction and planning of the national income of the GDR a complex mathematical model of the national economy has been developed in the division of national-economic accounts and analysis of the Central Statistical Office (CSO). By use of time series of the national in-come statistics the productive part of the economic reproduction cycle has been represented with its complicated subsystems, interactions and feed-backs.

Hans-Peter Höschel

### Statistical Methods for Supporting the Selection of Strategies in Automatization by Regard of Socio-Economic Factors

In introducing new technologies, not only technological, economical and organizational aspects should be in the focus of our considerations, but also the position of people within the labour process, a position which is effected by and effecting new technologies.

Sigrid Zeipert

### Model Analysis of Energy-Economy Interactions

The proper social and economic development of both individual countries, as well as of the whole world is impossible without the adequate supply of energy carriers. Energy is the indispensable for all branches of economic activity production factor. The manufacturing of fuel and energy is becoming more and more expensive because of the necessity to reach for more costly primary resources, application of more costly technologies, and also because of the increasing demands in environmental protection. The interactions between the energy system and the economy are becoming ever more complex. Therefore investigation of the development of energy system i.e. the complex of production, transport, distribution and utilization of energy, must be closely bound with investigation of the superior systea of the national economy.

Janusz Cofala, Krzysztof Ostromecki

### Cost Minimization in a Firm’s Power Station

Energy represents for every industrial site a fundamental process input which can only be substituted with great difficulty and without which industrial manufacture in general is impossible. The requirement is hereby concentrated on the utilization of distinct types of thermal energy for the process steam and heating, as well as for power and light. According to WOLF (1987) the total demand in 1981 for energy for industrial purposes in the Federal Republic of Germany was divided as follows — 75.9% for the manufacturing process, 10.9% for space heating and 13.2% for power and light. Although several energy resources were used equally for the production of heat, power and light were generated almost without exception by means of electricity. Heat as well as power and light, or rather electricity, are all commodities which have only a limited storage capacity. This leads to considerable procurement problems when the energy requirement fluctuates.

Günter Fandel, Joachim Reese

### Dynamic Energy Production Model

A dynamic one-cost timer double fixed cost model of producing and storing energy ist studied. An algorithm will be provided which generates optimal one-peak solutions, i.e. solutions satisfying that production is switched on at most one time.

Knut Richter

### The Growth of Electric Energy Production as an Example for the Application of the Evolon Model

The growth of electric energy production in different countries can be discribed by the evolon model during sufficiently undisturbed periods. The found cooperativity parameters of the model are in agreement with a classification rule, connecting these values with the character of the growth process. The data for some countries show decreasing absolute growth rates during the last decade. This leads to important changes in the growth dynamics of this key indicator of national economy. The evolon model reflects this increasing dominance of the saturation process with high accuracy. On the basis of this model an attempt can be made to project the electric energy production for the next decades. According to that model and in agreement with the data of some countries the increasing dominance of the saturation process characterized by falling absolute growth rates leads to a completly new situation in the growth dynamics in comparison to the situation in the last decades.

W. Mende, K.-F. Albrecht

### Analysis and Simulation of the Dynamic Behaviour of Saturated Steam Turbines of PWR Nuclear Power Plants

A digital simulation model of the components of a PWR (pressurized water reactor) type nuclear power plant has been developed in a modular structure in CRIP. This paper deals with the detailed modelling of the transient characteristics of turbines using saturated steam. This model calculates the discontinuity of the expansion line during transients simulating the turbine more exactly than by the conventional models.

Gacs Andras

### Optimal Power Control of a Nuclear Reactor by means of an Inverse Model

The control of nuclear reactor operation during instationary operating states is connected with high demands to process control in order to guarantee nuclear safety every time. Especially during start-up, a dynamic nonlinear process with power changes over several decades, the restriction of the reactor period is to be observed. In this connection problems of time-optimality have to be solved. Based on the point-kinetic equations an inverse reactor model has been derived. It permits to calculate such a control variable, the reactivity ς , which produces the wanted period T as the reactor output signal. By means of this inverse model a reactor period controller with a superposed power controller has been developed and realized by a microcomputer. The experiments with the zero-power reactor RAKE show that the controller is able to control both the reactor period and the power.

D. Hoppe, J. Götz

### Simulation Model and Microprocessor System for Digital Control of Hydraulic System in A Hydroelectric Power Plant

In this paper mathematical and corresponding simulation model for real hydroelectric power plant hydraulic system (lake, two headrace tunnels, three surge shafts and four pressure pipes) have been presented, together with microprocessor system which controls the observed system, and, if necessary, the generating unit so that the wave propagation does not exceed the, so called, critical level in the three surge shafts regardless of the requirements during the hydroelectric power plant operation. A similar microprocessor system (by the same author) has been in function in a similar hydroelectric power plant and has proved successful so far.

Milan Sodan, Seid Tesnjak, Ante Marusic, Tomislav Tomisa

### Power System Frequency Control Characteristics as a Function of Nuclear Power Plant Participation

When the participation of nuclear power plants in electric power system increases then they have to be expected to take an increasing part in the frequency and power control of the power system. However there are specific regulation problems in nuclear power plant control [2,3]. They cause, among others, a deterioration of power system frequency control characteristics.

Z. Domachowski

### Design of a Reduced Model for a Nuclear-Power Plant Under Earthquake Excitation

In this paper the nuclear-power plant will be discretized by using the finite element method. We get a system of ordinary differential equations. To reduce the simulation-time the original system will be approximated by a lower order system. For beginning we use in this paper a system of linear ordinary differential equations. In the next paper we extend the investigations also for nonlinear systems. The design of the reduced model will be carried out by using the method of minimizing the equation-error. The quality of the approximations will be examined by comparing the simulation results of the nonreduced model with the reduced model.

N. Model, S. Grossman, P. Dineva, L. Hadjikov

### Mathematical Modelling and Simulation Techniques of Complex Power Plant Pipe Network

The analysis of thermohydraulic transients is essential to plant design and safety. Phenomenological description of two phase flow is based on integral balance of mass, energy, and momentum DELHAYE [2]. The balance laws can be written for each control volume divided into subvolumes V1 and V2.

Frank Häusler, Gert-Steffen Rösel

### Structural Aspects of the (Natural and Artificial) Semantic Memory

The idea of semantic memory has proved to be useful both for interpreting psychological experimental data and as a key term of a computational technique for knowledge representation and information retrieval. The problem of the complexity of the structure of semantic networks and some qualitative aspects of kinetic models of the ontogenetic formation of semantic memory are discussed.

P. Érdi

### Estimation of Drug Need from the Health Projection — The Case of Insulin

The most important factors influencing drug need are disease prevalence and employed therapy. Our ability to predict future drug need is in most cases limited by the lack of adequate information on both these factors. One of the notable more favourable cases is diabetes though even here we are still far from the desired situation. A permanent rise in diabetes prevalence in most countries attracted much attention in recent years and epidemiological data about diabetes are gathered with a rapid pace. Even when these data are still rather imperfect, they can — together with the data about the demographic composition and characteristics of the studied population — form the necessary basis for building an epidemiological simulation model for the projection of the diabetic population. The second tiasic factor affecting the drug need is the employed therapy. Insulin therapy is indicated in a relatively well defined portion of diabetic patients and in Czechoslovakia the numbers of these patients are monitored by routine medical statistics.

František Hauser, Michal Anděl

### Numerical Simulation of Pulsatile Flow on Arteries

Since many years arterial blood flow has been systematically investigated, — both by experimental and theoretical methods. The contraction and relaxation of the left ventricle initiates a non-linear transient flow at the root of the aorta which results in complex wave patterns in the aorta. The waves are propagated throughout the arterial tree and reflected at major bifurcations, stenosis and other pathological phenomenas. So the characterization of pulsatile wave patterns is of particular interest to physiologists since irregularities in human circulation may be stated by analysing the time- dependent graphs of pressure and flow. Some of the related questions concerning the pulse wave propagation and the pulse wave reflection in arterial vessels may be answered by numerical simulations. Therefore a mathematical theory, based on an unsteady one-dimensional model of an incompressible fluid flow in a tapered,elastic tube is presented. The aortic valve is incorparated by special boundary conditions. The elasticity of the wall is described by an arctangent model cited from Langewouters et al. /3/. Wall friction is also taken into account as well as diffusion of fluid through the arterial wall. The hyperbolic model equations are discretizised in analogy to the implicit Beam — Warming finite differences method /8/ and numerically solved.

Thomas Pohl

### “Balance” Segments in Multicompartmental Models

Among the very important components of the General Method of Multicompartmental Models (GMMM — /l/) are so called segments. A segment is defined as a non-compartmental model approximation of the multicompartmental subsystem behaviour. One of the many possible types of segments are “balance” segments.

Milan Kotva

### Dynamoexpert: PC-Aided Simulation and Identification Tools for Biomedical Applications

The proposed system DYNAMOEXPERT is a set of powerfool and flexible tools for development and usage of mathematical models for various biomedical applications. Main concepts of mathematical modeling (simulation) are becoming being accepted now as a common research instrument in many areas including various bio-medical applications. Complexity of the whole human organism and its funcional (homeostatical and adaptational) sub-systems demands thorough development of problem- oriented data storing, processing and iterptetation methods. At the same time, available instrumental software does not seem to be satisfactory enough to supply regular everlasting support for large-scale long-term complex bio-medical research projects [1].

V. V. Shakin, A. A. Karasev, V. V. Przyjalkowski

### A New Model for Current Flow in Dendrites of Arbitrary Geometry

Current flow within dendrites or axons of nerve cells has usually been represented in the framework of the core conductor concept /2/. Starting with its standard assumptions, a new model of current flow in branching dendrites is derived. The method is a variant of the continuum approach employed in many different fields of research /1/. In the present case, the fine structure of the branching dendrites is replaced by an equivalent, locally homogeneous electrical medium described by averaged electrical quantities. For the mean potential and the mean current density in this medium partial differential equations of parabolic type are derived which represent natural generalizations of the simple case of a leaky cable.

A. Schierwagen, J. Francu

### Simulation System SONCHES Based Toxicokinetic Modelling as a Tool in Biological Monitoring

The use of an interactive simulation system (SONCHES) for toxicokinetic modelling will be described.The usual physiological based description of the toxicokinetics of inhaled gases/vapors (partition coefficient concept) is completed by an approach for urinary excretion of metabolites, water, creatinine. Various simulation experiments (scenarios:certain inhaled chemicals, different species — human, rat, variation in body build, physical exercise, and exposure pattern)are stored in a result library.The simulation system make a tool of the user’s model bank not only because of the facilities for post-processing of these results. Despite of the progress which has been made, further research is needed in the validation of our models.

G. Wünscher, V. Wenzel, M. Flechsig, E. Matthäus

### Parameter Estimation of the Blood Pressure Control System in Dogs

Experiments were carried out with trained conscious dogs and anaesthetized mongrels to identify the short-time regulation of blood pressure. The readings were used to establish parametric models in the form of discrete transfer functions, and the parameters were estimated by recursive least squares method and weighted recursive generalized least squares method.

Christian Fritzsch

### Computer-Graphic for Rotational and Mirror Symmetric Fractals with m Colours and a Determined Basic Design

The production of fractals 1 is controllable under geometric or aestetic points of view. In the sequence of conformal mappings {z_{j + 1}} = w({z_j}),(j = 1,2, \ldots ) the mapping function for a picture is unique represented by K-1 fixed points zFk and L poles zPL: w(z) = z + z\frac{{_{k = 2}^{_\prod ^K}(z - {z_{Fk}})}}{{{}_{l = 1}^{{}_\prod ^L}(z - {z_{Pl}})}} In this representations the source of the coordinate system is the fixed points zF1 = 0.

H. Schilling

### Simulation Environments

In the paper it is discussed that in systems simulation nowadays a simulator is becoming a synonym for a computer provided with a simulation environment, i.e. a programming environment dedicated to simulation. A simulation environment is defined as a programming environment of a computer, that is dedicated to systems simulation and that takes care for a flexible and intelligent interfacing between a user (i.e. the experimenter) and the system to be experimentally studied. The tasks a simulator has to support are depicted by a discussion about simulation studies in Research & Development. It is made clear that a simulation environment has to perform extra tasks in comparison to a normal programming environment of a computer: description, modelling, experimentation, knowledge handling, reporting. Several of these tasks ask for powerful and fast computers. An outline of a simulation environment is given. Implications with respect to hardware and software are discussed. It is argued that special-purpose simulators will be constructed as a general-purpose simulator provided with a special-purpose simulation environment rather than as specially hardware designed simulators.

L. Dekker

### Foundations of a General Process Theory

The concept of a process is often used in the various branches of mathematics, physics and system theory. Roughly speaking a process means a collection of time functions that in some sence belong together. Examples of processes are the set of all solutions of first order differential equations or sets of strings over an arbitrary set.

G. Wunsch

### Fast and Accurate Arithmetic for Vector Processors

In addition to the four elementary arithmetic operations, more advanced electronic computers such as vector and parallel computers often provide a number of compound operations as additional elementary operations. If pipelined compound operations like “multiply and add”, “accumulate”, and “multiply and accumulate” contribute essentially to the high speed of the system. Accuracy requirements lead to very similar operations. We identify a set of operations which meet both requirements: high speed and accuracy. After a brief discussion of implementation techniques for the simpler of these operations we present two methods and circuits which allow a fast and correct computation of the more complicated of these operations: “accumulate” and “multiply and accumulate”. The first method computes sums and dot products by making use of a matrix-shaped and pipelined arrangement of adders which cover the full floating-point range. The second method requires some local memory on the arithmetic unit. It permits a drastic reduction in the number of adders required. Both methods can also be used to build a fast arithmetic unit for microcomputers in VLSI technology.

Ulrich Kulisch

### Computer-Aided Systems Analysis: Methods and Tools

A survey about software tools for modelling, qualitative analysis, model aggregation and multi-criteria decision making at the ZKI is given. These software tools are developed in connection with industrial projects and combined in a unique framework for all components and a method bank for the steps of the analysis of complex dynamical systems up to the decision support for the control of these systems.

A. Sydow, K. Bellmann, E. Matthaeus, P. Rudolph, R. Śtraubel, A. Wittmuess

### A Flexible System of Precedence Coefficient Aggregation and Consensus Measurement

The paper presents a practical method for a software meant for the in-session processing of precedence data obtained from m judges and concerning n items. Precedence data of a judge can be given in a variety of ways — from a unique ordering to 1/2;-n (n-1) fuzzy precedence indices. First, the principles of processing are presented, then problems and suggested solutions. The latter refer to two notions: “expert agreement” as to the final precedence coefficients on the one hand and “resulting ordering” of the items on the other.

### Systematic Procedures to Derive Block Diagrams and State Equations from Bond Graphs

It is shown that a certain procedure can be followed to expand a causal bond graph [2] into a block diagram. Another procedure is presented to transform a bond graph into a set of equations which are ready for simulation and next into a set of state equations.

C. Breedveld

### Software Engineering Aspects of Computational Systems Analysis in Physics

An increasing role is assigned to computational systems analysis in applied physics.

P. Carl

### Set Deviations Modelling of Forest Fire Space State

In the paper the probability set methods assigned to set deviations estimation of the forest fire space random states in the function of their mean configuration form are considered. The deviation modelling methods as to union, interseption and as to mean deviation are suggested. The forest fire distribution identification algorithm as to average measure set and average measure deviation are presented.

T. N. Ivanilova

### An Attempt for an Adapted Model Generation Methodology: Example of Application in Building Energy Modelling

Issued only from intelligent beings, the conscient actions are based on prospective analyses and reasoning, and somewhere introduce what can be called modelling. In fact, these “intelligent” behaviours use an inductive processing: a conceptual internal representation is built and processed in order to predict the effects of different possible causes, the causes and their predicted effects have a direct interest for the intelligent being.

Louis Laret

### A Draft of a Component Model Library Manager, The “MODELOTHEQUE”: Analysis of an Approach, and First Results

Model users are generally not modelling experts. And modelling experts generally use field-specific language and implicit knowledge. Thus the users often prefer their old usual models, may be somewhat unaccurate, whose well known capabilities and limits allow easy results interpretation. This contribution is a proposal of a methodology which aims: 1. to produce a general and multifacetted documentation about the model, 2. to process this information for an aid to use the model in a component-based modelling approach.

Anne-Marie Dubois

### Models for Lake Eutrophication

Eutrophication is the answer of an ecosystem on increased nutrient loading. In the case of lake ecosystems it can threaten the ability of the lake for recreation or for using its water as drinking water. For meliorating the water quality the first step is the understanding the system to be modelled, the second step is forecasting its future in different nutrient loading conditions, revealing the response of the lake on certain measures. The first goal needs detailed models, large amount of state variables and processes, etc. while for fulfilling the second goal far simpler models are needed. At present the majority of models are deterministic, here one possible way of constructing stochastic model and its adventage is discussed.

Tibor Kutas

### Ways to Reduce Time Consumption of Statistical Modelling of Integrated Circuits by the Monte Carlo Method

Of the most widespread methods to determine parametric reliability the Monte Carlo method is the best to consider the specific character of integrated circuits (IC). Such common problems of statistical modelling as the inclusion of non-linearity of a circuit model, large number of random input variables and their statistical dependence for the Monte Carlo methods amount essentially to the problems of computation time and memory, i.e. time consumption.

A. I. Baškys, D. J. Zanevitčius, V. J. Gobis

### Mathematical Models of Non-Linear Elements in the Theory of Systems

Theory of systems is known to investigate objects of various physical nature and to base itself on mathematical simulation and a physical experiment. Methodologically the theory of systems is constructed in terms of the abstract theory of the complex systems in which there is a host of interrelated elements. Mathematical models of elements determine in many ways the system model.

D. Zanevičius

### The Medial Section as a Proportion for the Thermodynamical Equilibrium of Selforganizing Systems

A fundamental role in living systems is beeing played by the conception of thermodynamical equilibrium. In the paper /l/ the state of ther modynamical equilibrium is defined as such state of a system in which the value of the thermodynamical probability (and the entropy respectively) of the system remains unchanged under random fluctuations of the elements.

A. P. Stachov

### Recent Advances in Decision Technologies for Management

In this paper we provide an overview of the Decision Making Problems which face companies in a complex competitive environment. We then explore the technologies of expert systems, decision support systems and a hybrid between the two as potential tools for helping managers to tackle these decision making issues. Finally, a number of specific decision technology elements developed by the author are described and one of these is illustrated for helping with the solution of an important problem in Retail Banking i.e. determining interest rate mixes for savings and lending products in a competitive environment.