Filtered Jitter

Jitter is one popular way of generating samples for stochastic sampling in computer graphics. The Poisson disk distribution better approximates that of the human photomosaic. In this paper we examine the spatial and frequency space behaviour of a number of existing algorithms for generating stochastic samples and propose a new algorithm based on low pass filtering a jittered set of displacements. The distribution is at least as much like that of the human photomosaic as any existing algorithm, while being fast to compute.     

Versatile Tuning of Humanoid Agent Activity

In this paper, we present an integration framework for heterogeneous motion generators. The objective is to outline issues that are currently easily solved in professional post-processing systems used in film and game production but which cannot be transposed as is to real-time systems with autonomous agents. We summarise our approach for articulated agent-modelling and their animation by combining heterogeneous motion generators, such as real-time motion capturing, key-framing, inverse kinematics, procedural walking. We propose an agent/action-oriented framework. Activity properties such as action simultaneity and motion blending, spatial coherence, motion-flow update schemes, agent attachments, and location corrections, are the main topics handled by our generic animation framework. Numerous examples throughout the paper illustrate our approach and outline encountered problems and solutions or open research directions.     

Resource allocation in grid computing

Grid computing, in which a network of computers is integrated to create a very fast virtual computer, is becoming ever more prevalent. Examples include the TeraGrid and Planet-lab.org, as well as applications on the existing Internet that take advantage of unused computing and storage capacity of idle desktop machines, such as Kazaa, SETI@home, Climateprediction.net, and Einstein@home. Grid computing permits a network of computers to act as a very fast virtual computer. With many alternative computers available, each with varying extra capacity, and each of which may connect or disconnect from the grid at any time, it may make sense to send the same task to more than one computer. The application can then use the output of whichever computer finishes the task first. Thus, the important issue of the dynamic assignment of tasks to individual computers is complicated in grid computing by the option of assigning multiple copies of the same task to different computers. We show that under fairly mild and often reasonable conditions, maximizing task replication stochastically maximizes the number of task completions by any time. That is, it is better to do the same task on as many computers as possible, rather than assigning different tasks to individual computers. We show maximal task replication is optimal when tasks have identical size and processing times have a NWU (New Worse than Used; defined later) distribution. Computers may be heterogeneous and their speeds may vary randomly, as is the case in grid computing environments. We also show that maximal task replication, along with a c μ rule, stochastically maximizes the successful task completion process when task processing times are exponential and depend on both the task and computer, and tasks have different probabilities of completing successfully.     

Linear Programming Models for the User and System Optimal Dynamic Network Design Problem: Formulations,Comparisons and Extensions

In this paper we formulate a network design model in which the traffic flows satisfy dynamic user equilibrium conditions for a single destination. The model presented here incorporates the Cell Transmission Model (CTM); a traffic flow model capable of capturing shockwaves and link spillovers. Comparisons are made between the properties of the Dynamic User equilibrium Network Design Problem (DUE NDP) and an existing Dynamic System Optimal (DSO) NDP formulation. Both network design models have different objective functions with similar constraint sets which are linear and convex. Numerical demonstrations are made on multiple networks to demonstrate the efficacy of the model and demonstrate important differences between the DUE and DSO NDP approaches. In addition, the flexibility of the approach is demonstrated by extending the formulation to account for demand uncertainty. This is formulated as a stochastic programming problem and initial test results are demonstrated on test networks. It is observed that not accounting for demand uncertainty explicitly, provides sub-optimal solution to the DUE NDP problem.     

Experiments on the reliability of stochastic spiking neural P systems

In the area of membrane computing, time-freeness has been defined as the ability for a timed membrane system to produce always the same result, independently of the execution times associated to the rules. In this paper, we use a similar idea in the framework of spiking neural P systems, a model inspired by the structure and the functioning of neural cells. In particular, we introduce stochastic spiking neural P systems where the time of firing for an enabled spiking rule is probabilistically chosen and we investigate when, and how, these probabilities can influence the ability of the systems to simulate, in a reliable way, universal machines, such as register machines.     

Measuring observability by generalized information theoretic quantities

A normalized measure is established to provide the quantitative information about the degree of observability for the discrete-time, stochastically autonomous system. This measure is based on the generalized information theoretic quantities （generalized entropy, mutual information） of the system state and the observations, where the system state can be a discrete or a continuous random vector. Some important properties are presented. For the linear case, the explicit formula for the degree of observability is derived, and the equivalence between the proposed measure and the traditional rank condition is proved. The curves for the degree of observability are depicted in a simple example.     

Rényi information measure for a used item

Recently, in the literature, a measure of tail heaviness has been proposed based on Rényi entropy. This measure is very useful in the sense that it can be used to measure tail heaviness even for the distributions for which the kurtosis measure does not exist. Nadarajah and Zografos [Nadarajah, Zografos, Information Sciences 153 (2003), 119-138] have derived the expression for this measure for different univariate continuous distributions. But, this measure can only be used for the lifetime of a new item. In case of used item, this measure needs some modification. In this paper, we have modified the measure accordingly so that it can be used in the case of used item and also for the new item. We have also derived expressions for this measure for sixteen different univariate distributions and ten other standard distributions derived from the general distributions used in reliability and survival analysis. The most of the results obtained in the literature in this direction can be obtained as particular cases of our general results.     

A simulation optimization approach in production planning of failure prone manufacturing systems

In this paper, the implementation of a new method to control the production rate of manufacturing systems, based on the combination of stochastic optimal control theory, discrete event simulation, experimental design and response surface methodology is outlined. The system under study consists of several parallel machines, multiple-product manufacturing system. Machines are subject to failures and repairs and their capacity process is assumed to be a finite state Markov chain throughout the analytical control model. The problem is to choose the production rates so as to minimize the expected discounted cost of inventory/backlog over an infinite horizon. We first show that, for constant demand rates and exponential failure and repair times distributions of the machines, the hedging point policy is optimal. The structure of the hedging point policy is then parameterized by factors representing the thresholds of involved products. With such a policy, simulation experiments are combined to experimental design and response surface methodology to estimate the optimal control policy. We obtain that the hedging point policy is also applicable to a wide variety of complex problems including non-exponential failure and repair times distributions and random demand rates. Analytical solutions may not be easily obtained for such complex situations.     

客户机／服务器系统缓冲区容量的Petri网分析

本文在详细分析客户机／服务器架构的基础上,采用随机Petri网方法对客户机／服务器体系进行建模,着重就模型的缓冲区性能给出了相应的GSPN性能计算方法,并结合实例进行定性分析。     

Minimax games for stochastic systems subject to relative entropy uncertainty: applications to SDEs on Hilbert spaces

In this paper, we consider minimax games for stochastic uncertain systems with the pay-off being a nonlinear functional of the uncertain measure where the uncertainty is measured in terms of relative entropy between the uncertain and the nominal measure. The maximizing player is the uncertain measure, while the minimizer is the control which induces a nominal measure. Existence and uniqueness of minimax solutions are derived on suitable spaces of measures. Several examples are presented illustrating the results. Subsequently, the results are also applied to controlled stochastic differential equations on Hilbert spaces. Based on infinite dimensional extension of Girsanov’s measure transformation, martingale solutions are used in establishing existence and uniqueness of minimax strategies. Moreover, some basic properties of the relative entropy of measures on infinite dimensional spaces are presented and then applied to uncertain systems described by a stochastic differential inclusion on Hilbert space. An explicit expression for the worst case measure representing the maximizing player (adversary) is found.