Systematic Literature Review Of Particle Swarm Optimization Implementation For Time-Dependent Vehicle Routing Problem

Abstract

Time-dependent VRP (TDVRP) is one of the three VRP variants that have not been widely explored in research in the field of operational research, while Particle Swarm Optimization (PSO) is an optimization algorithm in the field of operational research that uses many variables in its application. There is much research conducted about TDVRP, but few of them discuss PSO's implementation. This article presented as a literature review which aimed to find a research gap about implementation of PSO to resolve TDVRP cases. The research was conducted in five stages. The first stage, a review protocol defined in the form of research questions and methods to perform the review. The second stage is references searching. The third stage is screening the search result. The fourth stage is extracting data from references based on research questions. The fifth stage is reporting the study literature results. The results obtained from the screening process were 37 eligible reference articles, from 172 search results articles. The results of extraction and analysis of 37 reference articles show that research on TDVRP discusses the duration of travel time between 2 locations. The route optimization parameter is determined from the cost of the trip, including the total distance traveled, the total travel time, the number of routes, and the number used vehicles. The datasets that are used in research consist of 2 types, real-world datasets and simulation datasets. Solomon Benchmark is a simulation dataset that is widely used in the case of TDVRP. Research on PSO in the TDVRP case is dominated by the discussion of modifications to determine random values of PSO variables.