Showing papers on "Multi-objective optimization published in 1982"

Multiobjective analysis of multireservoir operations

Abstract: The objective of the study reported herein is to develop practical procedures for the analysis of a multiple purpose, multiple facility reservoir system to guide real time decisions concerning the optimal operation of the system. Application is made to the California Central Valley Project (CVP). The five purposes (benefits), treated as objectives here in the multiobjective optimization, include (1) hydropower production, (2) fish protection, (3) water quality maintenance, (4) water supply, and (5) recreation. The constraint method is used to develop the trade-offs while a specially modified linear programing and dynamic programing algorithm is used for optimization. Noninferior sets can be obtained with each benefit parameterized singly and in various combinations. Two sets of monthly historical streamflows, one set corresponding to a drought year and the other set to an excess water year, are used to develop the corresponding noninferior sets. These procedures provide guidance for allocating the total benefits derived from a region's water resources and for operating the available system within all statutory, contractual, and other applicable constraints. A very high degree of decomposition of the typically large multiple purpose, multiple facility operation problem is made possible by the above technique, resulting in a rapid delineation of the noninferior policy set. The decision maker participates at various stages of the analysis and can request more or less detail with regard to the noninferior set. In our opinion, information is best presented to him via a series of two dimensional plots representing various cross sections of the noninferior set. Tabular presentations are not conducive to a good appreciation of the consequences of alternative benefit allocation policies.

Interactive algorithm for multiobjective optimization

Abstract: A man-machine interactive algorithm is given for solving multiobjective optimization problems involving one decision maker. The algorithm, a modification of the Frank-Wolfe steepest ascent method, gives at each iteration a significant freedom and ease for the decision-maker's self-expression, and requires a minimal information on his local estimate of the steepest-ascent direction. The convergence of the iterative algorithm is proved under natural assumptions on the convergence and stability of the basic Frank-Wolfe algorithm.

Existence and characterization of efficient decisions with respect to cones

Abstract: Existence and characterization theorems for the efficient (nondominated) set of decisions inR n are presented. The existence is proved when the set of decisions satisfies some compactness conditions. The efficient set is characterized in terms of the exposed efficient decisions when certain convexity and compactness conditions are satisfied.

Interactive Multiobjective Optimization by the Sequential Proxy Optimization Technique (SPOT)

Abstract: A new interactive, multiobjective, decision-making technique for obtaining the preferred solution of the decision maker from among a Pareto optimal solution set is called the sequential proxy optimization technique, SPOT. A time-sharing computer program for this interactive procedures implements man-machine interactive procedures. This paper summarizes the theoretical framework of SPOT and applies it to reliability design of a series-parallel system. The details of SPOT and the interaction processes to solve an illustrative example are in a separately available Supplement.

Optimization of STEOR networks via Markov renewal programming

Abstract: We show how to treat the problems of time optimization and cost optimization of STEOR networks (GERT networks with only nodes of the “stochastic exclusive-or” type) within the scope of Markov decision processes and present the related dynamic programming techniques.

Some pathological configurations of noninferior set appearing in multicriteria optimization problems of chemical processes

Abstract: The noninferior set of multicriteria optimization problems may sometimes include some inactive subset in which a scalar-valued objective function of weighted-mean type cannot be minimal. It is shown that there exists such an inactive noninferior subset within the noninferior set of two multicriteria optimization problems concerning a cross-current multistage extraction process and a fluidized reactor-heater system.

Generating Efficient Alternatives for Development in the Chemical Industry

Abstract: Industrial development can be seen as the process of changing the production structure by means of investment over the course of time. To control this development to the benefit of society while maintaining the profitability of the industry, decision makers must learn how socioeconomic changes and market conditions affect the static and dynamic properties of the production structure. This paper reports on the progress of collaborative research into the design of tools which could help decision makers to control development in the chemical industry. The basic approach is to formulate a model of the equilibrium state of the industry or, in the case considered here, of a particular subsector of the industry. The development process is initially described by a static multiobjective optimization problem, from which a dynamic multiobjective optimization problem is then derived. An example illustrating the use of this method for the pesticide-producing sector is given. The optimization problem and method for controlling industrial development put forward in this paper were worked out as part of the research program on Growth Strategy Optimization Systems (GSOS), sponsored by the Ministry of the Chemical Industry in Poland. This program is actually carried out at the Institute for Control and Systems Engineering (ICSE), part of the Academy of Mining and Metallurgy (AMM) in Cracow. The multiobjective optimization method for generating efficient alternatives and the related software were developed by the System and Decision Sciences Area at IIASA. This collaborative research was carried out within the framework of the agreement on scientific cooperation cosigned by IIASA and the AMM in June 1980.

Control system design by multiobjective optimization

Abstract: We consider posing control system design as a multiobjective optimization problem and survey some of the tools for solving this problem.

Multicriteria Optimization of Air-to-Air Missiles

Abstract: The design of an air-to-air missile system involves trade-offs between conflicting measures of effectiveness. Significant measures of effectiveness include fuel requirement, interception time, launch boundaries, real time computation requirements and sensor requirements. This paper formulates the design problem as a multicriteria optimization problem. A direct solution to the multicriteria optimization problem is computationally intractable. A two-step approach is considered to reduce the amount of computation. Next, the methodology is applied to optimize the propulsion control of medium range air-to-air missiles with two pulse rocket motors. Simplifications are made to the method of approach to realize a control law which can be used to determine the interpulse intervals of a multiple pulse rocket motor powered-medium range air-to-air missile. Some preliminary results are shown in the paper.