Author : Ashish Kumar Nemani
Publisher :
ISBN 13 :
Total Pages : pages
Book Rating : 4.:/5 (587 download)
Book Synopsis Combinational Approaches to Solve Scheduling Problems by : Ashish Kumar Nemani
Download or read book Combinational Approaches to Solve Scheduling Problems written by Ashish Kumar Nemani and published by . This book was released on 2009 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt: ABSTRACT: In this dissertation, we discuss classical scheduling problems especially in transportation industries. All of these problems come from real life applications and are very critical from the academic as well as the financial point of view. Being NP-hard problems there doesn't exist any well-defined algorithm which can solve them efficiently with good running time, and thus rules of thumbs are still being followed in practice, with very myopic use of optimization procedures. These decisions are usually worth billions of dollars per year and even a slight improvement will have a significant economic impact. Through efforts described in this document, we try to develop some holistic approaches along with heuristics, to get efficient and effective results for these decision problems. This proposed work has the potential of implementation in commercial grade software. We first suggest some hybrid approaches to solve the intermodal load planning problem, which generate very effective solutions within minutes. In our second problem, subset-disjoint minimum cost cycle problem, we suggest several exact and heuristic approaches to find the minimum cost cycles which contains at most one node from any subset. These problems occur very often as a subproblem of other other combinatorial problems. In the third problem, location routing problem, we suggest column generation algorithm and show its effectiveness by doing experiments with benchmark problems. In our last problem, which is very critical in all railway industries, we propose several models based on based on mixed integer programming, heuristics, and other hybrid approaches. These algorithms show a significant improvement over the current practices.