1,721,180 research outputs found
How Do Personal Preferences Influence the Flow Dynamics in Networks?
Full text linkWe present a model that describes how past experience and available information may influence the behavior of bounded rational agents over a fixed network. This work can be framed in the literature on network performance and, in particular, on how such a performance could be defined in terms of effectiveness in coordinating flow dynamics on a fixed network structure as well as how individual agents’ preferences and behaviors may lead to different network performance. Specifically, our model describes the agents’ flow using a mean-field approach that considers path preference dynamics. Such a dynamic highlights the fact that the agents choose their path based on both the network congestion state and the observation of the decisions of the agents that have preceded them. We introduce the reader to a set of assumptions and an approach that can be used to prove the existence of a mean-field equilibrium over a suitable set of time-varying mass distributions, defined edge by edge in the network. Finally, we discuss the limitations and possible future developments of the model and its applications to organizational networks
Some experiences in model development foragent-based manufacturing scheduling and control simulation
No full textA new cost function to solve multi-attribute decision making problems with nonseparable attributes
No full textThe authors deal with multiattribute decision problems, assuming the existence of a utility/cost function, and the possibility of preferentially nonindependent attributes. The aim is to identify a cost fuction, whose structure satisfies the following three basic assumptions: (a) it should allow a certain degree of preferential independence among the attributes; (b) its identification should be through the use of an efficient algorithm; and (c) it should reflect the decision maker's behavior. An algorithm to determine the optimal cost function, when this function is positive semi-definitive quadratic or multilinear Hessian, is discussed
Dynamic Decomposition of the Real-Time Railway Traffic Management Problem
No full textIn a railway network, traffic is often perturbed and trains must be rerouted and rescheduled. Doing so in large networks is a challenging task, which has been tackled through various decomposition approaches in the literature. In this paper, we propose an algorithm for managing traffic considering dynamic problem decompositions. It is an asynchronous algorithm based on the decomposition of the problem considering at each time step the smallest possible portion of the network and subset of trains. We prove that this algorithm guarantees to find an overall feasible solution if it exists, for single track networks with passing loops
The cell as a decision making units
No full textEach living cell needs to solve a resource allocation problem, in which multiple inputs (uptake fluxes) and outputs (secretion fluxes) are the outcome of the stoichiometry of biochemical pathways and the regulation of metabolic enzymes. Quantifying the efficiency with which a cell solves this resource allocation problem constitutes a basic question in ``cellular economics".
In this paper, we propose the use of Data Envelopment Analysis to define multi-dimensional yields that can capture the multi-dimensional nature of cell input-output processes.
Data Envelopment Analysis, by treating cells as decision making units, enables also to introduce the concept of efficiency frontier that is both intimately connected to the shadow prices of flux balance analysis and useful to estimate the phenotypic phase space from experimental measurements of fluxes
Fair and Sparse Solutions in Network-Decentralized Flow Control
No full textWe proposed network-decentralized control strategies, in which each actuator can exclusively rely on local information, without knowing the network topology and the external input, ensuring that the flow asymptotically converges to the optimal one with respect to the p -norm. For 1 < p < ∞ , the flow converges to a unique constant optimal up∗. We show that the state converges to the optimal Lagrange multiplier of the optimization problem. Then, we consider networks where the flows are affected by unknown spontaneous dynamics and the buffers need to be driven exactly to a desired set-point. We propose a network-decentralized proportional-integral controller that achieves this goal along with asymptotic flow optimality; now it is the integral variable that converges to the optimal Lagrange multiplier. The extreme cases p=1 and p=∞ are of some interest since the former encourages sparsity of the solution while the latter promotes fairness. Unfortunately, for p=1 or p=∞ these strategies become discontinuous and lead to chattering of the flow, hence no optimality is achieved. We then show how to approximately achieve the goal as the limit for p 1 or p ∞.Green Open Access added to TU Delft Institutional Repository 'You share, we take care!' - Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.Team Tamas Keviczk
Conjugate Direction Methods and Polarity for Quadratic Hypersurfaces
Full text linkWe use some results from polarity theory to recast several geometric properties of
Conjugate Gradient-based methods, for the solution of nonsingular symmetric linear
systems. This approach allows us to pursue three main theoretical objectives. First, we
can provide a novel geometric perspective on the generation of conjugate directions, in
the context of positive definite systems. Second, we can extend the above geometric
perspective to treat the generation of conjugate directions for handling indefinite linear
systems. Third, by exploiting the geometric insight suggested by polarity theory, we
can easily study the possible degeneracy (pivot breakdown) of Conjugate Gradient-
based methods on indefinite linear systems. In particular, we prove that the
degeneracy of the standard Conjugate Gradient on nonsingular indefinite linear
systems can occur only once in the execution of the Conjugate Gradient
A linear quadratic control problem with mean field dependent fixed costs
No full textThe paper deals with an optimal control problem
in one dimension, having affine dynamics, and a running cost
which is discontinuous in the control variable. More precisely,
besides terms which are quadratic in the state and control
variables, a mean field dependent fixed cost is paid as long
as the control is activated (not null). This leads to a problem
that, although simple, does not seem to fall into a known class.
Moreover, the outcome is completely different according to the
magnitude of the fixed cost in comparison to other parameters,
such as the (constant) disturbance appearing into the state
equation. By means of some tools of Bellman’s Dynamic
Programming and viscosity solutions, we are able to provide an
explicit formula for the value function in all different cases, as
well as a feedback formula for the optimal control, leading in
some subcases to chattering optimal controls. Finally a mean
field game is introduced where a set of homogeneous players
minimize each the cost functional of the single player control
problem, sharing the fixed cost. For such a game the existence
of equilibrium points is proved and their characterization is
displayed
- …
