Model and Algorithms for Scheduling Uniform Parallel Machines

Production scheduling helps businesses become more productive and efficiently satisfy consumer needs. Because it falls within the combinatorial category, scheduling work on uniform parallel computers to minimize the makespan measure is a difficult problem under single machine scheduling. The goal of this essay is to show how to design an effective algorithm for this particular situation.

The introduction chapter of this text covers single machine scheduling with a single machine, single machine scheduling with parallel machines, and research gap of unform parallel machine scheduling.

The goals of the challenge, the assumptions behind the scheduling of uniform parallel machines, and the plan for developing the algorithms are all presented in Chapter 2, which describes the problem. A model and an example of the model utilizing a small size problem are provided in Chapter 3 on the mathematical model for uniform parallel machines scheduling to minimize makespan. The simulated annealing algorithm to minimize makespan in the single machine scheduling problem with uniform parallel machines is presented in Chapter 4. It includes an overview of the algorithm, its skeleton, three newly developed simulated annealing algorithms, and a comparison of them to determine which is the best.

The development of a GA-based strategy to reduce the uniform parallel machine scheduling makespan is presented in Chapter 5. This covers the basic structure of the genetic algorithm, factors that impact the GA algorithm, techniques for assigning tasks to machines, and crossover strategies. Additionally, it entails the creation of four GA-based heuristics to reduce makespan and a comparison of them to determine which GA-based heuristic is optimum for scheduling jobs on uniform parallel computers.

In order to determine which is optimal, the best GA-based heuristic and the best simulated annealing technique are compared in Chapter 6 to minimize the makespan of scheduling jobs on uniform parallel computers. Appendix 1 for replication 1 and Appendix 2 for replication 2 contain the data needed to test the heuristics in Chapter 4, Chater 5, and Chapter 6.

The author expresses gratitude to all of his university colleagues and academicians for their support in releasing this work. The writing style of this text is simple to read.

P. Senthilkumar, B.E., M.E., Ph.D