摘要:

A stochastic learning automaton model based on relative reward strength is proposed for solving the job scheduling problem in distributed computing systems. The scheduling approach belongs to the category of distributed algorithms. An automaton scheduler is used for each local host in the computer network to make the decision whether to accept the incoming job or transfer it to another server. The learning scheme proposed makes use of the most recent reward to each action provided by the environment. This feature means that the automaton has the capability to handle a class of uncertainty such as workload variation or incomplete system state information. Simulation results demonstrate that the performance of the proposed scheduling approach is not degraded in the case of a change in workload and is better than the approaches of Fixed Scheduling Discipline and Joining the Shortest Queue under incomplete system information.

ISSN:0020-7721    

DOI:10.1080/00207729308949555

出版商:Taylor & Francis Group    

年代:1993

数据来源: Taylor

摘要:

This work considers the integration of quality and production responsibilities, and uses both quality and quantity as contracting attributes. The worker is better off with this arrangement not only through the increased compensation, but also through the job enlargement. In general the optimal contracts are strictly increasing in quantity and quality produced. Detailed comparative statics are derived for linearly additive production and quality-enhancing processes and power utility functions. It is shown that reductions in uncertainty and unit production cost, and increases in quality mark-up price and job-enlargement factors are encouraged by the principal.

ISSN:0020-7721    

DOI:10.1080/00207729308949556

出版商:Taylor & Francis Group    

年代:1993

数据来源: Taylor

摘要:

A new optimal control model of chemotherapeutic cancer treatment is presented including both the number of cancer cells and the level of toxicity as state variables. It is shown that under certain conditions, periodic drug administration is optimal. Although the model is highly stylized and too simple to describe realistic chemotherapeutic treatment patterns, it may be seen as a first step to prove that periodic treatment strategies are superior to therapies with monotonic decreasing or increasing dosage rates.

ISSN:0020-7721    

DOI:10.1080/00207729308949557

出版商:Taylor & Francis Group    

年代:1993

数据来源: Taylor

摘要:

A basic problem faced by organizations importing bulk commodities from abroad is when and how much to order so that the costs incurred will be minimized. Usually the demand is quite stable, yet the shipping time is unpredictable. An inventory model is built to consider importing different commodities from different regions to one destination. A constraint of limited storage space is imposed to meet real conditions. Based on the special structure of the problem, an efficient solution method is devised. To demonstrate the validity of the proposed model and the devised solution method, the coal shipping problem of the Taiwan Power Company (TPC) is formulated and solved. The results are very practical and are very convincing to the top management of TPC.

ISSN:0020-7721    

DOI:10.1080/00207729308949558

出版商:Taylor & Francis Group    

年代:1993

数据来源: Taylor

摘要:

Some elements of the systems approach to the analysis of an information flow in a dynamic environment are proposed. The resulting control system is designed to enable a manager to structure an information exchange taking into account the effects of delay. It matches very frequent real-life situations in which the following should be answered: ‘What is better, complete information but heavily delayed or incomplete information less delayed?’ The paper is primarily concerned with the mathematical representation of the functioning of autonomous groups, i.e. groups which themselves determine information requirements, flow and processing. Emphasis is placed on the role of delay in an evaluation of information flow.

ISSN:0020-7721    

DOI:10.1080/00207729308949559

出版商:Taylor & Francis Group    

年代:1993

数据来源: Taylor

摘要:

Uncertainty in dynamical systems may arise due to inaccuracies in modelling, parameter variations and external disturbances. As a result of this uncertainty, the performance index of an optimal control system deviates from its optimum value, which is referred to as the deterioration. A technique is presented to find a bound on the deteriorated performance index of optimum linear systems subject to bounded uncertainty. Uncertainty is incorporated as a forcing term in the system equations. To find the deteriorated performance bound, the performance index subject to the uncertain system is to be maximized within a specified time interval. The interchange theorem is used to interchange the maximizing and integral operations in the performance index functional to obtain a pointwise problem. Then, a Lyapunov technique, used to find reachable sets of uncertain systems, is applied to find the pointwise maximum values. The method serves as a measure of performance robustness or a measure of sensitivity to uncertainties. The problem is analytically solved for first-order systems. Finally, examples concerning first- and second-order systems are given as applications of the technique.

ISSN:0020-7721    

DOI:10.1080/00207729308949560

出版商:Taylor & Francis Group    

年代:1993

数据来源: Taylor

摘要:

The complexity of the dynamic behaviour in non-linear systems depends on the number of state variables involved and on how they are connected. In recent times the description of chaotic behaviour and of strange attractors have been the subject of numerous investigations. Most methods developed for the characterization of dynamic systems are based on the phase space representation. In the approach presented here, however, a system's behaviour is described in terms of the network comprising the interactions among the state variables. Information theoretical methods are introduced to quantify a network's organization along the dimensions of time and of the sign of the interactions. Examples are provided by applying the information measures to a simple network model and to the interaction network derived for the Lorenz equations.

ISSN:0020-7721    

DOI:10.1080/00207729308949561

出版商:Taylor & Francis Group    

年代:1993

数据来源: Taylor

摘要:

A bound is derived for the norm of the difference between the solution of a linear differential equation with known input, and an approximate solution obtained by expansion in a Chebyshev series. The approach is to derive a bound on the difference between two approximate solutions of different order. This bound is obtained in terms of the order of the approximation of lowest degree, the difference in the orders of approximation, and certain matrix norms. Extending this expression so that it is valid for arbitrarily large differences in order yields a norm bound on the difference between the approximate solution and the exact solution. The basic approach can be extended to other orthogonal function sets.

ISSN:0020-7721    

DOI:10.1080/00207729308949562

出版商:Taylor & Francis Group    

年代:1993

数据来源: Taylor

摘要:

This paper is concerned with multivariable adaptive control for a class of nonlinear stochastic systems with backlash inputs. A simple weighted adaptive control algorithm is presented for the control of such systems whose linear part possesses an arbitrary interactor matrix and is not necessarily minimum phase. The global convergence of the algorithm is established. Simulation results show that the algorithm has good control performances.

ISSN:0020-7721    

DOI:10.1080/00207729308949563

出版商:Taylor & Francis Group    

年代:1993

数据来源: Taylor

摘要:

An infinite horizon optimal economic growth model with consumption and non-convex technologies is studied. For this model, the existence of an optimal program is established and it is also shown that the system is in asymptotic equilibrium.

ISSN:0020-7721    

DOI:10.1080/00207729308949564

出版商:Taylor & Francis Group    

年代:1993

数据来源: Taylor