This paper proposes a modelling methodology that enables models to be created about the relationships between direct workers, their working environment and their subsequent performance. This methodology has been demonstrated and validated in a study at Found Motor Company. This study is described along with the simplifications and assumptions that were made to enable such experimentation.
Packet switching is an effective technology for integrating voice and data in a single network. An example for such a kind of a network is Internet. New service is emerging, and it is a Voice over Internet Protocol (VoIP). An important aspect of packet-switched voice is the reconstruction of a continuous stream of speech from the set of packets that arrive at the destination DTE. Because of the statistical nature of packet switching, each packet can encounter a different amount of delay in traversing a packet network. The problem described in this paper is how to design an appropriate voice reconstruction strategy that reproduces acceptable quality speech from packets that arrive with varying delay. The basic idea proposed in the paper is to measure a delay jitter between several points on the Internet. Our voice traffic generator, implemented on a workstation, sends packets that are 20ms long. Receiver measures interarrival time, keeps track of mixed sequences and lost packets. The measurement results are given in a form of average delay jitter. Receiver also keeps a record of all relevant packet information in a text file that is used as an input to a delay jitter predictor. Traffic prediction element predicts talkspurt characteristics and directly influences playout delay. For a prediction element, we propose Adaptive Neuro-Fuzzy Inference System (ANFIS).
This article presents a proposal for TMN management traffic characterization, followed by traffic generation models. The purpose is to offer mechanisms for the performance evaluation of such networks.
The paper deals with throughput time reduction in an emerging service system in self-service restaurants: the free flow configuration. Two modelling approaches, one an approximate and simple queueing model, the other a more accurate but elaborate discrete event simulation model, are described and applied to a real case. Special attention is given to the practical aspects of modelling and the practical value of both approximate and more accurate modelling. Aspects that influence throughput time performance addressed in this study include: bottleneck issues, layout of the free flow, service modes, capacity allocation, crowding and time dependency of the arrival pattern.
The dominant problems in ATM networks are network control and network resource allocation. An efficient resource allocation to each traffic source is an extremely complex task. Network must be managed in real-time in order to avoid congestion. Network congestion requires admission control that in turn requires traffic measurements. In this paper we propose two methods, one for the link capacity allocation, an the other for the estimation of a cell loss rate in an ATM switch that is based on the convolution of probability density functions of rate signals at the switch's input. Both methods are simple and can be easily implemented in a DSP in an ATM node. They also point to a possible application of discrete Fourier transform (DFT) to link capacity allocation and buffer design in high-speed networks.
This paper presents a new method to generate a milling tool path by means of path expansion considering a,chip volume machined. Compared with the previous studies of tool path generation that are known as the pattern projection, the proposed method is featured in terms of the path development or expansion. As preconditions, a ball-nosed end-milling tool is assumed and a surface to be machined can be represented in a geometric model of spatial lattice point. In the method, successive passing points of the tool path on the offset surface of machined surface are generated as a result of the promising tool center position. In the procedures determining the tool center position, every direction is examined as a candidate of the promising tool path direction with calculating a chip volume and the most suitable direction is selected of which chip volume is the nearest to the criteria index. If the expansion of the tool path failures, the back track of the path is done and the branch of the path expansion is tried. As an experiment, a trophy cup shape was tested and the milling tool path was generated. Finally the conclusions are briefly mentioned too.
Analysing cluster aggregates is an important problem in clonogenic assays. This measurement has to be carried out on 3D populations and, as far as possible, automatically. The method described in this paper is based on a 3D reconstruction of some serial slices. Different analyses techniques are used to measure cell organisation in 2D images. An efficient reconstruction method for this kind of problem is proposed, and tests for characterising cell distribution in a 3D medium are made. These preliminary results are discussed, showing the possibility of solving the initial problem using this kind of technique.
In this paper, the switched diversity methods are investigated for two types of wave propagation environment (urban areas and hilly areas). In this method signal from only one of the antennas is used at any one time. The theoretical performance limits of this type of space diversity are also computed. A linear predictor is designed which gave good estimates of the signal value at some future time. Finally the simulation results are obtained by using the linear predictor. From the BER (bit error ratio) results calculated the predictor switched diversity method has good improvements compared with the simple switched diversity method, especially at high speeds of receiver.
We propose a framework within which a discrete event simulation (DES) approach is used to model the impact of various screening policies for Alzheimer's disease (AD). A conceptual model is developed that incorporates aspects of patient demography, disease progression, and AD management services. The conceptual model is implemented using the Micro Saint DES software. Development of the framework, and choice of simulation software, permit the modelling of realistic scenarios, and evaluation of screening policies in terms of the total life years of a cohort. Estimates of system parameters are obtained from data in the literature. Validation of the model is established by comparison of AD prevalence obtained by simulation, when there is no screening programme, with data drawn from the literature. Preliminary results indicate that a screening programme for elderly people aged 75+ is only marginally inferior to one in which all individuals aged 65+ are screened. We propose that the framework developed herein could be adapted for evaluating screening policies for other chronic diseases, and extended to incorporate other evaluation criteria.
This paper examines the car following theory from the cybernetics perspective. It presents a fuzzy based approach for estimating the main parameters of the basic car following model. It maintains the structure of the classical car following model, limits its complexity provides means to take into account the main parameters which influence the behavior of the driver, while allowing for the formal verification of local and global stability of the stimulated traffic flow.
The influence of data parameters (sensor error, unexplained variance, sampling density and data distribution) on spatial data prediction quality is considered through the use of a spatial data simulator. Performance of linear and non-linear regression models (feedforward neural networks) is compared on simulated agricultural data, but the results can be generalized to geological, oceanographic and other spatial domains. For a highly non-linear response variable, non-linear models are shown to perform better regardless of unexplained variance and sensor error, but linear models outperform non-linear models when the sampling density of spatial data is not sufficient to produce accurate interpolated values. In the presence of non-homogenous data distributions, a significant prediction quality improvement can be achieved by using specialized local models assuming that distributions are properly discovered.
A method is described how to integrate robot control software into the standard mechanical simulator ADAMS(TM). The main point is that the robot designer can plug-in his micro controller routines "as is" into the simulation environment. We assume here that any relevant sensor input can be modeled up to a certain accuracy within the simulator. We implemented this approach to support the design of a walking robot called Scorpion*. First simulation results are presented concerning walking instabilities.
Structural modelling framework provides methods and tools for knowledge acquisition, representation and processing for diagnosis task solving, Knowledge captured in structural models are used to support several kinds of reasoning. In this paper a causal reasoning supported by the extended model of functional structure (FSM) is discussed The FSM is decomposed into a set of event trees. Causal relationships an evaluated qualitatively or qauntitatively and this information is processed using event tree simulation techniques for analysis of system's operation under critical events (faults). To carry out the analysis a piece of knowledge represented by an event tree is transformed into a corresponding muse-consequence rule. A deep knowledge rule base supports widely known forward and backward chaining that are vehicles for reasoning.
The partial buffer sharing scheme is very efficient in congestion control of multiple loss priority class traffic. In such a system, it is necessary to determine multiple buffer thresholds so that cell loss requirements can be satisfied. It is also very useful to define an optimal position of multiple loss thresholds. This paper presents an optimization algorithm for searching optimal position of multiple loss thresholds for the fixed load and for admissible maximal load. It also covers analytical simulation results of four-priority class buffer controlled by partial buffer sharing scheme.
Most engineering systems are characterised by a large number of measurement variables. When using Fuzzy Inductive Reasoning (FIR) methodology to model a complex system, a problem of computation time arises. The model can theoretically be computed, but, in practice, the computational cost is too high. If a decomposition of the complex system into subsystems can be achieved, the computational cost will be reduced: given a k-variables system, the cost of computing a global model, when using an exponential complexity methodology as FIR, is much higher than computing p models of j<k variables. In this paper a method to determine subsystems (subsets of related variables) of a complex system is proposed. The method is based on linear and non-linear statistical analysis. In order to show how the proposed methodology works, it is applied to a real data set which describes a garbage incinerator process, and an assessment on the obtained results is achieved.
Simulation study of logistics processes at the Baltic Container Terminal of the Riga Commercial Port is presented. The overall simulation system incorporates two types of models, operating within a common environment: Arena-based model and SLX-based model. Design of the serving both models, is discussed. Using the Arena-based model for evaluating terminal layout and equipment productivity is considered.
The present paper follows a work dealing with wheel/rail contact constraints published by Samin and Fisette in 1994 [1] and cam/follower contact constraints published in 1999 [2]. According to the latter publications, mechanical system subjected to contact constraints is modelled using the multibody approach and relative coordinates. The proposed methodology is applied to the vehicle dynamics, for which the wheel/ground contact is fundamental. In a lot of cases, one can admit that some contacting elements are significantly elastic: this is especially obvious in the contact region. In spite of this, the rigid contact model still has an important meaning and raison d'etre for pedagogic purpose but also for the analysis of rolling systems for which the contact slip is sufficiently negligible to justify pure rolling conditions.The goal of the present work is to describe the permanent contact between a wheel and the ground in the normal, tangent and longitudinal directions. The proposed form of the constraints equations is not homogenous: whereas the holonomic normal contact constraint has a purely geometrical nature, the tangent and the longitudinal contact constraint can only be expressed as kinematic non holonomic constraints. As a consequence, a "mixed" integration scheme involving the first and second order differential equations becomes necessary for the system analyse. The illustrative example will be the well-known bicycle which, although being a very familiar system, is a very interesting and fruitful application in vehicle dynamics.
The approach to the convergence of Finite Elements based Neural Networks (FE-based NN) is proposed after the introduction of its architecture and algorithms with a variational formulation. The approach consists of the energy function formulation and the projection method from which the range of values of the descent rate is derived. In order to apply the FE-based NN to dynamic problems, the size of the time step and the weights of the derivative of the unknown variable in the algorithm of the FE-based NN are discussed. The simulations of the assembly of parameters are carried out. The main results of the simulations are presented.
Terminals are an important part of any transportation system. It is in a terminal that the reallocation of goods flows occurs. Apart from the obvious handling of physical flows information handling is an important part of a logistic process. It both controls and reports the status of the logic system. It is necessary to have an integrated view of both the physical flows and the information flows when modelling logistic systems. In this paper some considerations that have to be made carrying out this modelling are discussed. A tool for port terminal modelling that incorporates information handling is also discussed. The tool is based on an extended version of combinatorial graph theory. Work carried out shows that a tool of this kind is useful for modelling terminals and the network approach makes the tool fairly easy to understand.
An object oriented simulation program is presented which allows calculating the time-dependent mass transfer in extracorporeal artificial organ systems. Every part of the extracorporeal circuit is modelled by a proper C++ class. Mass transfer is described by the one-dimensional advection-diffusion equation, which is solved with a split operator technique. This program is employed to study the influence of pulsatile pumping as in single cannula haemodialysis systems on the artificial kidney efficiency.
