Various stochastic programming problems can be formulated as problems of optimization of an expected value function. Quite often the corresponding expectation function cannot be computed exactly and should be approxim...
详细信息
ISBN:
(纸本)0780373073
Various stochastic programming problems can be formulated as problems of optimization of an expected value function. Quite often the corresponding expectation function cannot be computed exactly and should be approximated, say by Monte Carlo sampling methods. In fact, in many practical applications, Monte Carlo simulation is the only reasonable way of estimating the expectation function. In this talk we discuss converges properties of the sample average approximation (SAA) approach to stochastic programming. We argue that the SAA method is easily implementable and can be surprisingly efficient for some classes of stochastic programming problems.
We propose a new genetic algorithm to solve complex stochastic programming problems, in which possible combinations of parameters are provided as scenarios. The algorithm finds a solution efficiently using a statistic...
详细信息
We propose a new genetic algorithm to solve complex stochastic programming problems, in which possible combinations of parameters are provided as scenarios. The algorithm finds a solution efficiently using a statistical selection mechanism in addition to a sampling approach. In the algorithm, to reduce the computational demand, individuals are evaluated based on mean fitness in some scenarios sampled at random. Furthermore, to limit the probability that good individuals are excluded from the population by sampling error, selection in the algorithm is carried out based on statistical theory (i.e., Welch's test). Our approach significantly reduces computing time required to find high quality solutions for stochastic facility location problems.
In recent years, tools for solving optimization problems have become widely available through the integration of optimization software (or solvers) with all major spreadsheet packages. These solvers are highly effecti...
详细信息
In recent years, tools for solving optimization problems have become widely available through the integration of optimization software (or solvers) with all major spreadsheet packages. These solvers are highly effective on traditional linear programming (LP) problems with known, deterministic parameters. However, thoughtful analysts may rightly question the quality and robustness of optimal solutions to problems where point estimates are substituted for model parameters that are stochastic in nature. Additionally, while many LP problems implicitly involve multiple objectives, current spreadsheet solvers provide no convenient facility for dealing with more than one objective. This paper introduces a decision support methodology for identifying robust solutions to LP problems involving stochastic parameters and multiple criteria using spreadsheets. (C) 2002 Elsevier Science B.V. All rights reserved.
We investigate the sensitivity to tax change of multi-stage portfolio allocation over a discrete time investment horizon. Special taxation rules within wrappers grouped a number of risky assets are integrated with mul...
详细信息
The electric power industry is undergoing restructuring and deregulation. We need to incorporate the uncertainty of electric power demand or power generators into the unit commitment problem. The unit commitment probl...
详细信息
Many optimization problems can be expressed naturally in a recursive manner. Problems with a dynamic structure are commonly expressed in this way especially when they are to be solved by dynamic programming. Many stoc...
详细信息
Many optimization problems can be expressed naturally in a recursive manner. Problems with a dynamic structure are commonly expressed in this way especially when they are to be solved by dynamic programming. Many stochastic linear programming problems have an underlying Markov structure, and for these a recursive definition is natural. Real-world examples of such problems are often so large that it is not practical to solve them without a modeling language. No existing algebraic modeling language provides a natural way of specifying a model using a dynamic programming recurrence. This paper describes the advantages of recursive model definition for stochastic linear programming problems, and presents the language constructs necessary to implement this within an algebraic modeling language.
Agricultural nitrate emissions within a river catchment are, due to rainfall and other sources of natural variation, uncertain. A regulator aiming to reduce nitrate emissions into surface and groundwater faces a trade...
详细信息
Agricultural nitrate emissions within a river catchment are, due to rainfall and other sources of natural variation, uncertain. A regulator aiming to reduce nitrate emissions into surface and groundwater faces a trade-off between reliability in achieving emission standards and the cost of compliance to agriculture. This paper explores this trade-off by comparing different assumptions about the probability distribution of nitrate emissions and thus the probabilistic constraint included in the catchment model. Three categories of probabilistic constraints are considered: (1) nonparametric, (2) normal and (3) lognormal. The results indicate that the restrictiveness of the non-parametric assumption could lead to a significant reduction in profit relative to the normal and lognormal. The lognormal assumption, although it is theoretically correct, cannot be generalised to the case of correlated emissions. However, ignoring the dependence between different sources of nitrate emissions introduces more bias than mis-specifying their distribution. Therefore a probabilistic constraint based on a correlated normal distribution of emissions gives the best approximation for nitrate emissions in this study. (C) 2002 Elsevier Science B.V. All rights reserved.
The paper gives an overview over the broad area of distributed decision making (DDM). In achieving a systematic procedure a general framework is developed that allows to describe the numerous approaches in DDM in a un...
详细信息
The paper gives an overview over the broad area of distributed decision making (DDM). In achieving a systematic procedure a general framework is developed that allows to describe the numerous approaches in DDM in a unified way. Focusing on application areas the paper is not only considering various fields in the management sciences, like hierarchical production planning, supply chain management, or managerial accounting, but is regarding other disciplines as well. Particularly, economics and computer sciences are investigated as to their specific contributions to DDM. It turns out that each field and discipline elaborate different aspects of DDM which particularly for OR could be used to solve concrete highly involved DDM problems. (C) 2002 Elsevier Science B.V. All rights reserved.
作者：
Martel, AUniv Laval
Fac Sci Adm Ctr Technol Org Reseau Quebec City PQ G1K 7P4 Canada
This paper develops rolling planning horizon policies to manage material flows in multi-echelon supply - distribution networks with relatively general stochastic demand processes and procurement, transportation, inven...
详细信息
This paper develops rolling planning horizon policies to manage material flows in multi-echelon supply - distribution networks with relatively general stochastic demand processes and procurement, transportation, inventory and shortage cost structures. Initially, the problem is formulated as a stochastic program with recourse, and its deterministic equivalent program is approximated by a multi-echelon lot-sizing model based on "risk inflated effective demands." A DRP - decomposition of this approximate model, which can be used with planning time fences or allocation algorithms, is then introduced. The use of expediting actions is also discussed. Finally, through a set of simulation experiments, the solutions obtained with our planning policies are compared with the solutions given by a classical DRP approach using safety stocks. The results show that the approach proposed leads to significant improvements.
We present a framework for solving some types of 0-1 multi-stage scheduling/planning problems under uncertainty in the objective function coefficients and the right-hand-side. A scenario analysis scheme with full reco...
详细信息
We present a framework for solving some types of 0-1 multi-stage scheduling/planning problems under uncertainty in the objective function coefficients and the right-hand-side. A scenario analysis scheme with full recourse is used. The solution offered for each scenario group at each stage takes into account all scenarios but without subordinating to any of them. The constraints are modelled by a splitting variables representation via scenarios. So, a 0-1 model for each scenario is considered plus the non-anticipativity constraints that equate the 0-1 variables from the same group of scenarios in each stage. The mathematical representation of the model is very amenable for the proposed framework to deal with the 0-1 character of the variables. A branch-and-fix coordination approach is introduced for coordinating the selection of the branching nodes and branching variables in the scenario subproblems to be jointly optimized. Some computational experience is reported for different types of problems. (C) 2002 Elsevier B.V. All rights reserved.
暂无评论