• Earthquakes and Structures
• /
• v.10 no.2
• /
• pp.429-450
• /
• 2016

An attempt has been made to incorporate the concept of collapse safety margin into the procedures proposed in the performance-based earthquake engineering (PBEE) framework for direct earthquake loss estimation, in which the collapse probability curve obtained from incremental dynamic analysis (IDA) is mathematically characterized with the S-type fitting model. The regressive collapse probability curve is then used to identify non-collapse cases and collapse cases. With the assumed lognormal probability distribution for non-collapse damage indexes, the expected direct earthquake loss ratio is calculated from the weighted average over several damage states for non-collapse cases. Collapse safety margin is shown to be strongly related with sustained damage endurance of structures. Such endurance exhibits a strong link with expected direct earthquake loss. The results from the case study on three concrete frames indicate that increase in cross section cannot always achieve a more desirable output of collapse safety margin and less direct earthquake loss. It is a more effective way to acquire wider collapse safety margin and less direct earthquake loss through proper enhancement of reinforcement in structural components. Interestingly, total expected direct earthquake loss ratio seems to be insensitive a change in cross section. It has demonstrated a consistent correlation with collapse safety margin. The results also indicates that, if direct economic loss is seriously concerned, it is of much significance to reduce the probability of occurrence of moderate and even severe damage, as well as the probability of structural collapse.

• Proceedings of the Korean Statistical Society Conference
• /
• 2005.05a
• /
• pp.203-207
• /
• 2005

A loss system where two types of customers arrive in accordance with two independent Poisson processes is considered. An efficient recursive formula is developed for calculating the loss probability when the number of servers is large. Some practical examples regarding the performance evaluation of telecommunications networks are discussed.

• Journal of applied mathematics & informatics
• /
• v.22 no.1_2
• /
• pp.273-283
• /
• 2006

We obtain an asymptotic behavior of the loss probability for the GI/PH/1/K queue as K tends to infinity when the traffic intensity p is strictly less than one. It is shown that the loss probability tends to 0 at a geometric rate and that the decay rate is related to the matrix generating function describing the service completions during an interarrival time.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.15 no.6
• /
• pp.2037-2047
• /
• 1991

Strength is not simply a single given value but rather is a statistical one with certain distribution functions. This is because it is affected by many unknown factors such as size, shape, stress distribution, and combined stresses. In this study, a model of loss probability is proposed in view of the fact that one of the fundamental configuration of nature is hexagonal, for example, the shapes of lattice unit, grain, and so on. The model sues the concept of loss of certain element in place of Jayatilaka-Trustrum's length and angle of cracks. Using this model, the loss probability due to each loss of certain elements is obtained. Then, the maximum principal stress is calculated by the finite element method at the centroid of the elements under the tensile load for the 4,095 models of analysis. Finally, the failure probability of the brittle materials is obtained by multiplying the loss probability by the ratio of the maximum principal stress to theoretical tensile strength. Comparison of the result of the Jayatilaka-Trustrum's model and the proposed model shows that the failure probabilities by the two methods are in good agreement. Further, it is shown that the parametric relationship of semi-crack lengths for various degrees of birittleness can be determined. Therefore, the analysis of the failure probability suing the proposed model is shown to be promising as a new method for the study of the failure probability of birttle materials.

• Korean Journal of Mathematics
• /
• v.11 no.1
• /
• pp.35-43
• /
• 2003

We study the weak convergence to Fractional Brownian motion and some examples with applications to traffic modeling. Finally, we get loss probability for queue-length distribution related to self-similar process.

• Journal of applied mathematics & informatics
• /
• v.24 no.1_2
• /
• pp.529-534
• /
• 2007

We obtain an explicit expression of the loss probability for the PR/M/1/K queue when the offered load is strictly less than one.

• Journal of KIISE:Information Networking
• /
• v.29 no.2
• /
• pp.181-187
• /
• 2002

Connection admission control(CAC), which decides whether or not to accept a new call request, is one of the most Important preventive congestion control techniques in asynchronous transfer mode(ATM) networks. To develop a practical CAC scheme, first we propose a "Modified Cell Loss Probability MP${\nu}"$, which is based on "Virtual Cell Loss Probability P${\nu}"$, taking into account mean burst duration of input traffic source and buffer size in ATM networks. MP${\nu}"$ computes more accurate cell loss probability than P${\nu}"$ without increasing computational complexity, since P${\nu}"$ is formulated simply form the maximum and the average cell rate of input traffic. P${\nu}"$ is overestimated as compared to the real cell loss probability when the mean burst duration is relatively small to the buffer capacity. Then, we Propose a CAC scheme, based on "Modified Virtual Bandwidth(MVB)" method, which may individualize the cell loss probabilities in heterogeneous traffic environments. For the proposed approach, we define the interference intensity to identify interferences between heterogeneous traffic sources and use it as well as MP${\nu}"$ to compute MVB. Our approach is well suitable for ATM networks since it provides high bandwidth utilization and guarantees simple and real time CAC computation for heterogeneous traffic environments.heterogeneous traffic environments.

• Proceedings of the Korean Operations and Management Science Society Conference
• /
• 1996.04a
• /
• pp.543-546
• /
• 1996

Recently, some research has been done to analyze the asymptotic behavior of queue length distribution in ATM (Asynchronous Transfer Mode) multiplexer. In this paper, we relate this asymptotic behavior with the asymptotic behavior of decreasing cell loss probability when the buffer capacity is increased. We find with reasonable assumptions that the asymptotic rate of queue length distribution is the same as that of decreasing cell loss probability. Even under different priority control schemes and traffic classes, we find that this asymptotic rate of the individual cell loss probability of each traffic class does not change. As a consequence, we propose the upper bound of cell loss probability of each traffic class when a priority control scheme is employed. This bound is computationally feasible in a real-time. The numerical examples will be provided to validate this finding.

• The Korean Journal of Applied Statistics
• /
• v.23 no.4
• /
• pp.747-757
• /
• 2010

This paper studies the Markov modeling of multiclass loss systems supporting several kinds of customers. The concept of unit for loss systems is introduced and the method of equal probability allocation among units is especially considered. Equilibrium equations and limiting distribution of the loss systems are studied and loss probabilities are computed. We analyze an example of a simple system to gain an insight about general systems.

• Journal of Korea Society of Industrial Information Systems
• /
• v.11 no.1
• /
• pp.1-6
• /
• 2006

Queueing models are good models for fragments of communication systems and networks, so their investigation is interesting for theory and applications. Theses queues may play an important role for the validation of different decomposition algorithms designed for investigating more general queueing networks. So, in this paper we illustrate that the batch size distribution affects the loss probability, which is the main performance measure of a finite buffer queues.