Algoritma Heuristik untuk Menyelesaikan Masalah Lintasan Terpendek Stokastik

Abstract

In this thesis, we study the shortest path problem with stochastic arc length. According to different decision criteria, propoced the concepts of expected shortest path, shortest path and the most shortest path, and present three new types of models: expected value model, chance-constrained programming and dependent chance programming. In order to solve these models, a hybrid intelligent algorithm integrating stochastic simulation and genetic algorithm.

Dalam tesis ini dibahas masalah lintasan terpendek melalui panjang busur stokastik. Berdasarkan kriteria keputusan yang berbeda dibahas konsep atau modelmodel dari expected shostest path, shostest path, dan most shostest path. Model nilai harapan (expected value model) program batas kemungkinan (chance constrained programming)dan program kemungkinan dependen (dependent chance programming), dapat diselesaikan dengan menggunakan algoritma hybrid serta simulasi stokastik dan algoritma genetik.