Discrete Invasive Weed Optimization Algorithm for Graph Based Combinatorial Road Network Management Problem

Abstract

Invasive Weed Optimization (IWO) Algorithm is a nature inspired swarm based continuous domain optimization meta-heuristics which mimicry the expansion-cum-survival strategy of the weeds in favorable, rich and unwanted regions which happens to be the best solution in terms of optimization with respect to competition, growth and nutrition. These unwanted plants are in consistent competition and opposition from the other members of the nature either directly or indirectly and as a result their way of living, foraging and sustaining are the most robust and challenging. This optimization technique has been proven to be successful in many continuous parameter domains due to their unique spreading characteristics and optimization search methods. In this work we have extended the utility of the invasive weed optimization algorithm for graph based combinatorial optimization for path search and planning for vehicle routing from a source to destination. The problem can be viewed as a multimodal optimization problem where selection of a certain sequence of multimodal solutions would be best solution. For this we have modified the classical IWO to suit the graph based situation and made necessary change in implications to cope up with the graph parameters. The convergence rate of the Discrete Invasive Weed Optimization (DIWO) Algorithm is being compared with Ant Colony Optimization (ACO) and Intelligent Water Drop (IWD) algorithm with an application on a road graph model for route optimization for vehicles with respect to multi-objective of travelling and waiting time.