A Multi Objective Scheduling Scenario in Industry Perspective Using Particle Swarm Optimization with Mutation Strategy

Abstract

Scheduling aims at allocation of resources to perform a group of tasks over a period in such a manner that some performance goals such as flow time, tardiness, lateness, and makespan can be minimized. To sustain in the current competitive environment, it is essential for the manufacturing firms to improve the schedule based on simultaneous optimization of performance measures such as makespan, flow time and tardiness. The problem of finding robust schedules (schedule performance does not deteriorate in disruption situation) or flexible schedules (schedules expected to perform well after some degree of modification when uncertain condition is encountered) is of utmost importance for real world applications as they operate in dynamic environments. newlinePSO is an effective algorithm which gives quality solutions in a reasonable computational time and requires less number parameters to be tuned in comparison to other evolutionary approaches. However, PSO has an inherent drawback of being trapped at local optimum due to large reduction in velocity values as iteration proceeds and poses difficulty in reaching at best solution. This drawback can be effectively addressed using Mutation, a commonly used operator in genetic algorithm, can be introduced in PSO so that premature convergence can be avoided. Logistic mapping can be used to generate chaotic numbers instead of random numbers to improve the solution diversity. newlineIn this dissertation work, a novel particle swarm optimization (PSO) algorithm has been proposed for solving the single objective as well as multi-objective scheduling for flexible flow shop and job shop scheduling problems. Methodology for obtaining robust schedule is proposed to deal with uncertain situation in flexible flow shop and job shop scheduling. It is demonstrated that solution quality improves when PSO algorithm is embedded with chaotic numbers and mutation. newline newline