Many of simulation based learning algorithms have been developed to obtain near optimal policies for Markov decision processes (MDPs) with large state space. However, most of them are for unichain problems. In view th...
详细信息
ISBN:
(纸本)9783642028939
Many of simulation based learning algorithms have been developed to obtain near optimal policies for Markov decision processes (MDPs) with large state space. However, most of them are for unichain problems. In view that some applications involve multichain processes and it is NP-hard to determine whether a MDP is unichain or not, it is desirable to obtain an algorithm that is applicable to multichain problems as well. This paper presents a rollout algorithm for multichain MDPs with average cost. Preliminary analysis of the estimation error and parameter settings are provided based on the problem structures, i.e., mixing time of transition matrix. Ordinal optimization and Optimal Computing Budget Allocation are also suggested to improve the efficiency of the algorithm.
Structured singular system, depending on a parametric vector are considered. The identification of the parameters is analyzed in terms of the input-output behavior of the system. The role of the reachability and obser...
详细信息
ISBN:
(纸本)9783642028939
Structured singular system, depending on a parametric vector are considered. The identification of the parameters is analyzed in terms of the input-output behavior of the system. The role of the reachability and observability properties in this analysis is studied and a characterization of the structural identifiability property is given. Finally, the structural identifiability of a positive reachable system is Studied.
We use orders of magnitudes of variables and parameters of a chemical system described by an ordinary differential equation, to obtain a partition of the state space in boxes (hyper-rectangles). From the fast system i...
详细信息
ISBN:
(纸本)9783642028939
We use orders of magnitudes of variables and parameters of a chemical system described by an ordinary differential equation, to obtain a partition of the state space in boxes (hyper-rectangles). From the fast system in each box, we derive rules of transition, and obtain a transition graph. This graph can be used for a qualitative simulation and validation of the system.
In this short paper we show how the convergence of the iterative aggregation-disaggregation methods for computing the Perron eigenvector of a large sparse irreducible stochastic matrix can be improved by an appropriat...
详细信息
ISBN:
(纸本)9783642028939
In this short paper we show how the convergence of the iterative aggregation-disaggregation methods for computing the Perron eigenvector of a large sparse irreducible stochastic matrix can be improved by an appropriate ordering of the data and by the choice of a basic iteration matrix. Some theoretical estimates are introduced and a fast algorithm is proposed for obtaining the desired ordering. Numerical examples arc presented.
Water Supply Systems (WSS) are clearly dynamical systems. Processes associated with WSS include design, planning, maintenance, control, management, rehabilitation, enlargement, etc. Modeling and simulation of these pr...
详细信息
ISBN:
(纸本)9783642028939
Water Supply Systems (WSS) are clearly dynamical systems. Processes associated with WSS include design, planning, maintenance, control, management, rehabilitation, enlargement, etc. Modeling and simulation of these processes can be performed by using a number of variables and constraints that are non-negative in nature. Demands, diameters of pipes, flowrates, minimum pressure at demand nodes, volume of reservoirs, are only a few examples, taken from the purely technical context. In this paper we will focus on the design of WSS. This a mixed discrete-continuous constrained optimization problem that is addressed here by the use of an evolutionary technique based on swarm intelligence. Robustness is enforced by adding reliability to the system both to cope with abnormal conditions and by considering the likelihood of different state and load conditions. Application to a real-world problem is also provided.
This contribution is a natural continuation of a series of papers devoted to analysis of models utilized by specialists in Cell Biology around E. Bob] and W. Boos. Our novelty may he seen in enriching the models in di...
详细信息
ISBN:
(纸本)9783642028939
This contribution is a natural continuation of a series of papers devoted to analysis of models utilized by specialists in Cell Biology around E. Bob] and W. Boos. Our novelty may he seen in enriching the models in direction of controllability in the spirit of biology engineering. Besides the standard properties of the models such as existence of appropriate solutions and their uniqueness the following issues are of interest: Asymptotic behavior (e.g. steady states and pseudo-steady states), controllability and also special features such as various types of symmetries, periodicity etc. Our aim is focused on periodicity of solutions of models whose state objects share the properties of concentrations, i.e. probabilities.
We present a voting system that is based on an iterative method that assigns a reputation to n + m items, a objects and m raters, applying sonic filter to the votes. Each racer evaluates a Subset of objects leading to...
详细信息
ISBN:
(纸本)9783642028939
We present a voting system that is based on an iterative method that assigns a reputation to n + m items, a objects and m raters, applying sonic filter to the votes. Each racer evaluates a Subset of objects leading to an n x in rating matrix with a given sparsity pattern. From this rating matrix a formula is defined for the reputation of raters and objects. We propose a natural and intuitive nonlinear formula and also provide an iterative algorithm that linearly converges to the unique vector oh reputations and this for any rating matrix. In contrast to classical outliers detection, no evaluation is discarded in this method but cash one is taken into account with different weights for the reputations of the objects. The complexity of one iteration step is linear in the number of evaluations, making our algorithm efficient for large data set.
The finite horizon Linear-Quadratic (LQ) optimal control problem with nonnegative state constraints (denoted by LQ(+)) is studied for positive linear systems in discrete time. Necessary and sufficient optimality condi...
详细信息
ISBN:
(纸本)9783642028939
The finite horizon Linear-Quadratic (LQ) optimal control problem with nonnegative state constraints (denoted by LQ(+)) is studied for positive linear systems in discrete time. Necessary and sufficient optimality conditions arc obtained by using the maximum principle. These conditions lead to a computational method for the solution of the LQ+ problem by means of a corresponding Hamiltonian system. In addition, necessary and sufficient conditions arc reported for the LQ(+)-optimal control to he given by the standard LQ-optimal state feedback law. Sufficient conditions are also reported for the positivity of the LQ-optimal closed-loop system. In particular, such conditions arc Obtained for the problem of minimal energy control with penalization of the final state. Moreover a positivity criterion for the LQ-optimal closed-loop system is derived for positive systems with a positively invertible (dynamics) generator
This paper considers the servomechanism problem for MIMO positive LTI systems. In particular, the servomechanism problem of nonnegative constant reference signals for stable MIMO positive LTI systems with unmeasurable...
详细信息
ISBN:
(纸本)9783642028939
This paper considers the servomechanism problem for MIMO positive LTI systems. In particular, the servomechanism problem of nonnegative constant reference signals for stable MIMO positive LTI systems with unmeasurable Unknown constant nonnegative disturbances under strictly nonnegative control inputs is solved using a clamping LQ regulator.
暂无评论