SIMULATION

The simulation process involves the collaboration of different participants, such as simulation analysts, programmers, statisticians, and users of the simulation software. Many simulation tasks such as modeling, verification, validation, and design for experimentation require the participants to meet. It is understood that these meetings are time-consuming and expensive. This paper proposes a collaborative environment to help with the tasks of discrete event simulation software development using the World Wide Web platform. The environment, named GroupSim, is based on a collaborative computer system and uses the concepts of distributed modeling with automatic program generation and distributed control of experimentation. The authors show some examples to illustrate the use of the environment and discuss some issues related to collaborative environments such as concurrency control, access control, awareness, and performance.

This article presents a stochastic model of lymphocyte recruitment in inflamed brain microvessels. Recent studies about the inflammatory process of the brain that leads to multiple sclerosis have revealed that lymphocyte extravasation is a sequence of dynamical states, mediated by partially overlapped interactions of different adhesion molecules and activation factors. This study’s model of lymphocyte recruitment is based on process algebras for mobile systems. The biochemical system is modelled as a set of concurrent processes of the biochemical stochastic π-calculus. Processes are driven by suitable probability distributions that quantitatively describe the rates and the times at which reactions to simulations occur. The results of the model reproduce, within the estimated experimental errors, the functional behavior of the data obtained from laboratory measurements.

Broadband packet networks must dynamically recognize some network flows, such as those that deal with disaster response, military operations, or emergencies, as having greater importance than others. An architecture of geographically distributed ticket servers has been proposed to issue importance tickets to identify the priority that a flow should be given in the current dynamic network context. Here the potential viability of such an architecture is studied with respect to the performance requirements of ticket servers, network overhead that would be added, and the severity of longer connection setup times. A hybrid simulation approach is used. Network and ticket server performance are modeled using prototype results combined with analytical techniques for networks of queues. Two scenarios are studied—a hurricane event and an office building bombing event—and ticket server performance requirements and connection setup delays were not found to be prohibitive to the successful implementation of such an architecture.

Robotic arms often perform industrial tasks requiring complex dextrous manipulation within constrained spaces. For example, automobile unibody assembly can require more than 5000 welds, with many performed within the vehicle’s interior. An arm can be designed specifically for this type of task by permuting link lengths and degrees of freedom (DOF) to find a set of feasible designs. Each design can be evaluated for joint-angle displacement, dexterity, simulated speed, and consumption of available redundancy. A heuristic search increases the probability of having the needed kinematic structure. Hyperredundant designs of up to 10 DOF can be created, and searches often yield minimized-DOF designs. The path-planning technique combines pseudo-inverse velocity control with the concept of attractive poles to allow maneuvering through complex enclosures while avoiding many obstacles. This research provides a means of rapid-prototyping robotic arms for enclosed spaces and can yield many designs locally optimized for given tasks and environments.