Genetic Algorithm Based on Idempotent Algebra Methods for RCPSP

Abstract

Resource-constrained project scheduling problem is an important practical task of production management. We consider the problem with resources in monetary form and a criterion for maximizing the net present value. The problem takes into account the technological relationships between activities. This paper presents the method of search for admissible schedules by solving linear equations over an idempotent semiring. In the genetic algorithm the vectors from an idempotent semimodule are used to describe the individuals of the population. Computational experiments demonstrate the efficiency of the genetic algorithm.