Process Modelling

Simulation is regarded as a strategic tool in the field of process development and process control. It is used to present certain aspects of an existing system, or a system to be developed, as models, or to reproduce them. It allows to examine systems which are too dangerous, too expensive or impossible to be tested. Precondition for simulation is the existence of a general or reduced process model which describes the important characteristics of the process with mathematical exactness. While in general process modelling it is in the foreground to collect and evaluate the knowledge on a certain process as completely as possible, in the reduced process model the boundaries of the process parameters are important. Such a process model has to be as exact as necessary within these parameters, while beyond the validity of these parameters exactness is of secondary interest. So as not to confine this introduction to process modelling to the technological, production-oriented area only, the introduced standardised term process model is now generalised as a brief outlook. A limitation in the area of nature is arbitrary and disadvantageous, hence one has to make an attempt to put process models in a systematic order in other areas, too.

Most important for a more sophisticated process modelling is to limit the involvement of users in the overall modelling process to the inclusion of existing a priori knowledge while making this process more automated and more objective. Automatic model generation is based on these demands and is a powerful way to generate models of ill-defined problems. “KnowledgeMiner” is a powerful and easy-to-use modelling tool designed to support the modelling process on a highly automated level and which presently implements three advanced self-organising modelling technologies: GMDH, self-organising fuzzymodelling and Analog Complexing. Keywords. Self-organising modelling, automatic model generation, GMDHalgorithm, neural networks, fuzzy modelling, Analog Complexing

Well defined procedures for the construction of models are necessary in order to support the real applicability of modelling techniques in the industrial settings where practical engineering means are required to model and evaluate complex systems and productive processes. This paper describes the application of two different approaches to the construction of Petri Nets models and provides some general hints on the modelling process. In particular, starting from a simple working example we suggest that a behavioural approach, modularity in the model construction phases, and integration of different modelling techniques could contribute to improve the efficiency in many fields of research, development and its controlling.

In this paper a high-level, user-friendly, yet executable process modelling language is introduced. The modelling of a process’ structure is supported through ER-like diagrams, whilst its behaviour can be specified in terms of state transition diagrams with guarded transitions, refined operations, and event handlers. It is enforced that process models consist of well separated process fragments which are subject to reuse. The enviroment supports the implantation of knowledge gathered during process execution into the process model to enhance assistance for the project management.

This paper describes how inexact processes as presented in sporting events can be recorded, analysed, and evaluated by means of neural networks and fuzzy modelling.

The computer ecological intellectual system (CEIS) is developed in the investigated regions on the basis of the object-orientated approach with the purpose to be used in systems of processing ecological information. (Voytenko, Zagorodni, 1997). The system has developed dialogue means, which allow setting base data maps for their processing through internal and external models of natural and agricultural processes. The result of such processing is new maps of the economic, ecological or social characteristics. The maps are created on the basis of researched region division into geographical elements for the supervision of necessary characteristics changes of dynamics. Thus, there is an opportunity to realise the analysis and forecasting. CEIS is logically and structurally subdivided into the subsystems. The subsystem is an abstraction. Its subclasses are versions of objects that represent the global characteristics of any mental or real territorial subsystem. According to such project and hierarchy of objects, it is obvious, that objects of derivative classes from the abstract generalised class and can submit the specialisation of the subsystems. Hence, at the centre of consideration -we have a subsystem, that displays a real current state of a pattern, which answers one of the characteristics, say, of current radiating loading in view of accommodation of pollution sources, natural-climatic conditions, sociologiceconomic structure and so on. The minimal specification of the project is determined as follows: - Biological subsystem (ground, microorganisms, viruses, plants, animals, birds); - Subsystem of environment (climatic conditions, pollution); - Subsystem of social development; - Subsystem of managing and economy. One of the external models, which is represented in CEIS, is a model of a haricot bean plant development in ontogenesis, nodule nitrogen fixing during a virus infection in different ecological conditions (Zagorodni, Boyko, Beiko 1995). As it is known, virus is a very sensitive system at change of the natural factors, therefore it can be used as the indicator of the environment state. The model uses the initial information as maps of the environment subsystem of CEIS (temperature, humidity, PH of soil, radiating, pestisidal and industrial soiling) and is a basis for the construction of new maps - maps of infection, productivity and its quality, pollution of a crop and so on. Also, through such a model, it is possible to find optimum conditions for the cultivation of a healthy vegetative material. The processes, which pass in vegetative organism, are separately given in the model. For each process, it is constructed of a function of potential force and a function of real force, which is defined under conditions of action of the laws of conservation of substances and energy. Thus, CEIS can be applied in the research of ecological processes in different conditions of investigated regions. Then the region is broken into geographical elements with the given set of abstract subsystems. As an example, a model of development of the plants in different conditions of Ukraine is shown.

Due to necessary specifications of manifold co-ordination mechanisms for managing resource inter dependencies and corresponding control data flows, the design of simulation models describing business processes often tend to generate considerably high complexity itself With regard to economic efficiency of process modelling and simulation studies it is essential to cope with this complexity on a systematic basis. The paper will take up this problem, which has so far rarely been regarded. Coming from the Guidelines of Modelling (GoM), a project of the University of Munster, three major mutually dependent topics are focused upon, and are applied in a case study: View-orientated phases of the model construction, the use of model components and the application of modelling conventions.

In order to support a more structured process design, many businesses today are designing their business processes using methods and tools for process modelling. Structured process design and planning increases the likelihood that newly designed processes can be realised in a timely manner, because process participants understand the need for the changes and understand how to integrate them with existing processes and resources.

Performance Measures and performance measurement systems are the basis of quality and quantity-related generation. They control business objects such as organisational units and business processes. The enhancement of current process modelling methods with possibilities to model performance measures allows companies to create a prospective company controlling system. Therefore, it is necessary to visualise the strategic objectives linked with the key strategic measures and integrated with the performance measures of the business objects. This paper points out the use of process modelling extended with strategic and operational performance measurement for all types of organisations.

Design and improvement of manufacturing systems is a very complex and creative planning task In Small and Medium Enterprises (SMEs), it is carried out in the course of planning projects by a group of enterprise planners who belong to different departments but work in close cooperation. Each of the team members has already been supported by sophisticated modelling and simulation tools to work on a specialist part of the model of the manufacturing system. Influences and effects between the modelling activities of different planners arise dynamically during the planning process due to the creative nature of planning tasks. They are caused by logical inter dependencies and overlapping data that exist between the partial planning models and specialist working fields of the planning experts. Transparency of all those modelling activities of planners that effect other collaboration partners is important for successful collaboration. An IT-enabled support of the group awareness and an intelligent reaction to mutually interrelated modelling activities is presented here. It is an essential part of a groupware-based toolbox for cooperative modelling, besides a concept for cooperative problem solving already developed and implemented in a prototype. Handling, controlling and coordinating the interdependent modelling activities between planning experts mainly depends on the planning issue, the characteristics of the planning project and the course of the whole modelling process. As planning issues are unique, requirements on a computer-based planning environment will change from one project to another. For example, the cooperation partners in a project group and their specific skills, the modelling tools used and the inter dependencies existing between the specialist working fields may change. To realise cooperative modelling in SMEs, it is necessary to enable an adaptation of the groupware-based toolbox to changing organisational, information and technical demands.

Static business process models (for example, flowcharts), as well as dynamic simulations, have been amongst the most widely used and effective methods for studying and analysing business processes. However, the transition from time-independent static descriptions of business processes to time-dependent dynamic simulation models may present modellers and decision-makers with a variety of problems, mainly related to data collection and experimentation. In this paper, we present a real-life case study of business process modelling. We illustrate the modelling process followed and the various challenges faced by modellers when enriching the static process models with the numerical data needed to drive the simulation runs. We present a number of specific issues that needed the attention of modellers, as well as some potential measures that can be taken to alleviate such problems. Drawing on the findings of our study, we discuss a novel approach for integrating models at different levels of abstraction to support the transition from static to dynamic process models.

In modern companies, the use of large information systems is increasing, in order to support the phases of the product development process. The Institute of Production Technology and Quality Management and the Institute of Technical and Business Information Systems have worked together on an integration of software systems at various levels, in order to support the engineers at the development of products. The aim of this is the support and automation of the process of preparation of offers for castings with focus on modelling and execution of the identified processes. This article describes a process sequencial control, based on the PACO-Integration framework, which is explain ad using the example of the work scheduling in foundries and the submission of proposals of a dynamic reaction of the process control on varying conditions.

Even though many workflow management systems (WfMS) are commercially available, only few existing systems are supporting business processes which are of main interest for the enterprise. This paper deals with the requirements of business process modelling and workflow modelling. These activities are not yet sufficiently supported by common methods. Based on the requirements, problem areas are identified and illustrated by SAP’s Business Workflow (SAP BWF). Finally, we suggest some solutions to meet the outlined problems.

Proceeding from the requirements placed on business process models and the evaluation of present-day methods and tools, this contribution presents a procedure for flexible process management. The prime focus is on the flexibility and especially the practicability of such a procedure. In order to master the complexity of operational systems, these aspects have been examined in various models in accordance with diverse criteria. In the concept presented here, these models received the necessary integration to allow flexible administration of both allocated tasks and the use of process information.

This paper describes a method of process benchmarking to support the dynamic modelling of business processes using a decentralised organised service industry of the insurance industry as an example. The way in which the integration of business process management and service system formation contribute to complexity control and reduction is shown. One of the possibilities for complexity management in service companies is the approach taken by the benchmark-based dynamic business process management. Through the integrated approach of service systems formations, the company is able to offer competitive customeroriented services.

Globalisation and rapidly changing business environments require enterprises which are able to identify success potentials of the future quickly and to develop them on time. “Strategic Production Management is a comprehensive management methodology developed to overcome these obvious challenges. The method OMEGA has been developed at the Heinz Nixdorf Institut for designing business processes. It is possible to transfer process and product data out of the engineered modelling-tool PRESTIGE into EDM- or Workflow Management Systems. Furthermore, a new approach for the visualisation of business processes is presented. Virtual Reality Technology is used to experience a simultanous view on the physical model of the factory as well as the business process model.

This paper describes the development of a system for the support of end users in making complex decisions. After surveying the according application area, stable variants (as processes, SQL statements, and programmes) are modelled in terms of components in the spirit of componentware. In addition, robust mechanisms for combining components are offered to the end user. Thus, the user may compile (new) solution strategies while being guided by the system. The flexible exception handling is guaranteed by the concepts of the componentware and by combining and adapting given components by the user at run time.

This paper describes a model to demonstrate a track bound line operation in public transport systems. After a short overview on the basic elements of the track bound line operation, the mathematical model is presented. The following simulation shows some typical scenarios of disturbances. Finally the essential conclusions for transport operators are drawn.

The contribution demonstrates the possibility to establish the Hamiltonian of an electrical system in two forms, dependent on a covariant- and contravariant, description of these systems, respectively to establish the canonical equations of motion. A basis system will be created for an electrical circuit, which is embedded in the Euclidean space.

Based on the equivalence of current-carrying circuit-loop and magnetic dipole density the field of a current-carrying air-core coil is approximated by the field of two magnetic equivalent charges which are to be attached to the winding axis at the distance of the coil ends. Thereafter the influence of a highly permeable core on the equivalent charges of the air-core coil are indicated with a known approximation equation, then the final equivalent source model of an El-shaped small transformer is developed. In the final example the maximum stray field of a small transformer, which - as expected - occurs along the winding axis, is estimated and compared to concrete measured values.

Process models in the energy supply system are frequently used to mathematically delineate the interrelation between components (i.e. consumers, generating plants and heat distribution systems) within this system. Various process models can be deployed depending on the type of approach in solving contentious questions in order to explain by example, the system assumed in the energy distribution system. As far as plant upgrading is concerned, a specific model will have to be developed to illustrate on the one hand the flow of energy between individual components within the system, and one which allows for a definite regulation of the expenses incurred from the installation of generating equipment and distribution systems on the other. It is imperative that certain procedures be simplified, go that a computer-aided solution to inherent problems can be found. The process model described here allows us to simultaneously compare all possible energy supply options by computer therefore, the best economic and ecological option can be ascertained. This process model thus represents one alternative to the method deployed for embracing particular variants which allows for only one option to be computed at any one time.

The following paper describes different methods of the investigation of numerical models for hydrodynamic couplings. First an overview of the methodology of modelling is presented. Analytic physical, black-box and hybrid modelling are explained in detail The experimental identification of the models for the hydrodynamic coupling is presented. Different modelling methods and the appendant results are compared. A critical reconsideration of the methods and the results are presented at the end of the paper.

Actually, there is a trend of integrating simulation based components in enterprise information and decision support systems. The available data in production planning and control software systems can be used to construct the models, the objective-oriented realisation of simulation experiments and the economic evaluation of the results. In this context, the conception and realisation of integrated systems is very important for the analysis, modelling, simulation and optimisation of business processes. The paper describes an approach of the production scheduling problem under consideration of high complex process conditions. The new control strategies are based on hierarchical concepts. Local and global mechanisms are used for the order scheduling. These methods are integrated in a software system for short time production planning and control.

In manufacturing technology the term prototyping is usually known in the context of rapid prototyping, the rapid building of physical product models. Usage of this term in the context of process modelling is new and unusual The concept of process prototyping aims to provide a tool, that enables the user to build up models of manufacturing processes for tests and studies of relevant parameters. From the large number of methods for simulation of manufacturing processes the most suitable have to be selected. Criteria for this selection are shown and explained. As an example, a prototype of the laser welding process, realised in a finite element system, is discussed. Possibilities opened up by this model are evaluated and further improvements are pointed out.

This paper describes a decision-support tool for the design of new mills and the modernisation of older ones, as well as the operational control of foundries. The modular structure and the flexible handling of this tool enables rapid generation, simulation, and evaluation of a number of different plant variants. The tool described here is currently under development. Individual components, such as the flowsheet simulator, have already been developed and are now being used to meet the challenges outlined below.

The production of brown coal briquette bundles in a completely automated production plant imposes strict quality requirements on the pressing of the briquettes. Thus the packaging of bundles which consist of several layers of briquettes requires that each layer has approximately the same length. As the pressing of briquettes with exactly predefined measurements cannot be mastered technically, a part of the production must be picked out with a suitable quality surveillance procedure. The quality surveillance procedure that was used in production could be mathematically described exactly and was subject to an indepth analysis. It could be verified that, besides the question for the optimal choice of single parameters, it is essential for this particular application that the performance of a quality surveillance procedure does not depend on the dynamic behaviour of the factors of production. The output of the factory could be increased considerably by eliminating these problems and optimising the production plant. Besides the mathematical modelling, this result was based on the optimisation and simulation by a computer-based model.

In micro abrasive blasting (MAB), a fine abrasive powder is accelerated in an air stream. When the powder hits the target with a velocity near the velocity of sound, material is abraded from the target surface. In this paper it is shown for the first time that, by comparing experiment and theory, the abrasion mechanism during micro abrasive blasting can be described mathematically. By investigating the effective mechanism for blasting of brittle materials and the concentration distribution of the blast medium in the jet of compressed air a process model for micro abrasive blasting can be established. With only few experimentally measured values it is possible to apply the procedure for computer simulation of profile forming structuring techniques.

For illustrating the injection moulding process a partial neural network structure is developed which takes into account the relationships between machine settings, process-sequence variables and quality parameters. The model is actualised continuously by an adaptation mechanism. It predicts the quality of the product which is defined individually by a quality vector and corresponding specification limits. During production the predicted quality vector is compared with the calculated intervention limits for each quality parameter. If a quality value lies out of its intervention interval the actual quality deviation is calculated and passed on to the controller. The relationship between changes of certain machine settings and its qualitative effects on the quality values must be accurately known for controller design. This relationship can be ascertained through an experimental plan or empirical experience of the machine operators. The design of experiments technique is applied in the first examination of moulds. The resulting data contains the influences of varied machine settings on the defined quality values. The collected data, changes of manipulated values and corresponding quality value changes, can be fixed in an expert system. This expert system can be extended with empirical experience know-how of the process which the machine operators have. Alternatively both kinds of knowledge could be integrated in a fuzzy module which is able to calculate the necessary manipulations of the machine settings. First experiments show that this hybrid design is able to compensate disturbances on quality parameters as batch variations to a certain degree.

The use of Computational Fluid Dynamics in the design of a pharmaceutical processor is described. A simple computational model of the process was developed using a single phase flow. The results of this preliminary study showed good agreement with experimental tests. The model was further developed to consider the effect of solid particles on the processes occurring within a single vessel pharmaceutical processor. Having obtained converged solutions the model was used to optimise the design of the processor. The design optimisation was done on the basis of comparisons with conventional pharmaceutical processing equipment. Optimisation of the process parameters are important as they are known to affect the functionality of the final product.

To control and monitor pumping wells, drainage systems in mines and excavations pits, deposits and industrial or agriculture contaminants, it is necessary to know the processes in the soil and groundwater zone. There are physical, chemical and/or biological processes. The physical processes are described mathematically by a system of partial differential equations (PDEs). The groundwater flow described by the heat-conduction equation and the transport of substances by the convection-diffusion equation. For a few of the simplified groundwater relationships one can use simple equations. For the decision support of the water experts and the simulation of a few important processes in the soil and groundwater zone a Computer Aided Engineering (CAE)-Groundwater system is developed at the Dresden University of Technology.

This paper aims to contribute to a better understanding of simulation applications. To concretise the topic, modelling approaches of the spreading of a contamination plume in groundwater, the radial symmetric flow towards an auger hole and an activated carbon fdter simulation are presented. Common to all examples is the flow property - the porous medium.

Since 1925 salty sewage water of the potash-industry (Werra-region, Middle Germany) has been deposited in the geological setup. A quasi-closed system was assumed. It was found out that a part of the salt water ascend to the surface again. Time series-analysis allow a precise analysis of the outcoming salt water. To separate trend and periodical shares the time series were filtered numerically. The Fourier-transformation allows an investigation of the periodic shares. After the evaluation of the data it seems possible that two bodies of water united into one single body of waste water. The spreading of salty water in the underground was modelled by the concept of cellular machines.

A theoretical treatment of miscible phase concept and its threedimensional realisation will be presented. As a practical background, the simulation of water-, air- and vapour migration around waste repository in the subsurface, under special consideration of the impact of an additional thermal gradient servers as a second driving force. One approach is to handle hazardous gases that may be the result of the burial of such materials beneath thick layers of compacted clay, where the clay’s small pores may act as a diffusive barrier that retards the release of harmful gases to the atmosphere. The mass migration is considered to take place in the water phase, air phase as well as in the vapour phase. The concept of condensation/evaporation is used to interconnect the vapour and water phase. Gaseous flux through the subsurface low pervious material cannot be described exactly by Fick’s law. Applying Hirschfelder’s approach for dilute gases the diffusion operator was divided into three components, the binary diffusion, the Knudsen diffusion, and the thermal diffusion, which regards the impact of the thermal gradient. In contrast to the mass migration, heat migration is considered in four phases among which migration through the solid matrix plays a dominant role. Both of the migration equations were interconnected by the terms of relative permeability, partial pressure, viscosity and moisture contents.

The paper will dicuss some statements about the structural analysis of dams published in other papers. The main groups of problems concerning the adequate modelling of the dam structure, analysed in more detail are: development of new methods and possibilities for their application (informatics, theory of probabilities, fuzzy set theory); the system approach in the structural analysis of dams; physical and mathematical modelling of the investigated structure: effort required, role of the active Codes; hierarchical levels in the structural modelling of dams; problem solving with insufficient input information (floods and earthquakes) ; validation, verification and calibration of the structural model; uncertainty modelling in the structural analysis of dams.

Interrelations between variables of freshwater quality were investigated by time series analysis methods. Especially wavelet transformations are used to identify the dynamic ecosystem behaviour. The basic idea is to analyse ecological processes according to time and scale. The original ecological signal can be represented in terms of wavelet expansions while data operations can be performed by the corresponding wavelet coefficients. Data taken from investigations of sediments of hypertrophic riverine lakes showed different interrelations between water temperature and soluble reactive phosphorus of pore water under aerobic and anaerobic conditions. Each time if the temperature of the upper sediment layer is higher than the temperature of the pelagic water the phosphate content of the overlying water increases and vice versa. Therefore, one topic of wavelet analysis was the qualitative identification of the water temperature-phosphate-sediment subsystem.

A simple and portable model which is based on real field situation and is comparable compatibility to the laboratory core sample model, the modified pressure infiltrometer technique, is introduced for the determination of fieldsaturated hydraulic conductivity, an essential parameter of the vadose zone. The field -saturated hydraulic conductivity, KFS, is obtained from measurements of the steady flow rates. The field-saturated hydraulic conductivity is also obtained by this device using the falling head principle at the site of granite soil. The device is also used for field core sampling on which laboratory constant head, as well as falling head tests, are carried out. The field device is validated through comparisons to laboratory core sample experiments and other existing methods. This paper describes for the first time a versatile field device, which can be used for laboratory modelling and representing good performance for in situ determination of hydraulic parameters in a short time.

To allow spatial query and spatial production i.e. analysis in an Geographic Information System (GIS) environment a rule-based approach is suggested. It enables analysis based on spatial conditions of maps and additionally non-spatial facts. The rule base documents the spatial modeling process and can be reused for similar routine decisions. Examples form the delineation of well head protection areas (WHPA) illustrate this approach. The language for the rules is Lisp extended by spatial analysis functionality using a spatial vector structure (Open GIS Standard or ESRI shape files). It is implemented in Java.