Resource Allocation by Convex Monotonous Programming Application to an Otec Energy Production Platform

• Published in: Modeling, identification and control Vol. 4; no. 1; pp. 25 - 31
• Main Authors: Mjelde, Kåre M., Fjeld, S.
• Format: Journal Article
• Language: English
• Published: Kristiansand Norsk Forening for Automatisering (NFA) 1983; Norwegian Society of Automatic Control
• Subjects: convex programming; non-linear programming; Resource allocation
• ISSN: 0332-7353; 1890-1328
• DOI: 10.4173/mic.1983.1.2

A problem is considered of the minimization of a convex and monotonically increasing cost function subject to an upper bound of a convex and monotonically decreasing failure probability function. An algorithm is defined for the parametrical solution of the problem with respect to the selected failure probability level; the algorithm is applicable to problems with cost and failure probability functions that are not necessarily separable, thus extending a previously given algorithm for separable functions. A condition is given for the relaxation of the convexity assumption on the functions, and it is observed that the problem of the minimization of the failure probability subject to an upper bound on the total cost is solved simultaneously with the given problem. An anlytically solvable numerical example is given and an application to the optimization of an OTEC energy production platform is briefly described, with reference to a complete description; it is concluded that the given extension of the algorithm to non-convex functions is important in this application.