Randomized Local Search for Real-Life Inventory Routing

Abstract

International audienc

An incremental approach using local-search heuristic for inventory routing problem in industrial gases

“…We highlight few important papers that deal with a similar problem. Aghezzaf et al (2006), Archetti et al (2007), Campbell and Savelsbergh (2004), Chien et al (1989), Yu et al (2008), and Benoist et al (2011) all study a multi-period IRP where the decisions are carried out over a finite horizon. We should note that most of the studies dealing with the infinite horizon problem use a distribution policy that is similar to the fixed partition policy (first introduced in Anily and Federgruen, 1993), direct deliveries, order-up-to level policy and zero-inventory ordering (Bertazzi et al, 2002;Chan et al, 1998).…”

“…More recent research papers that deal with optimal distribution of industrial gases supply-chains are by You et al (2011), Benoist et al (2011), Ellis et al (2014 and Marchetti et al (2014). You et al (2011) focus on strategic decisions with long-term planning horizon using decomposition and continuous approximation approaches.…”

“…Marchetti et al (2014) assess the benefits of optimal coordination of production and distribution in industrial gas supply-chains. Benoist et al (2011) propose a method to solve a real-life IRP for operational scheduling using a randomized local-search heuristic for short-term planning horizon. This method uses a novel surrogate objective function based on long-term lower bounds and report savings exceeding 20% on average compared to solutions built by expert logistics analysts.…”

“…• The randomized local-search heuristic considered by Benoist et al (2011) is improved by using a new long-term objective function. • A novel incremental approach is proposed to optimize the IRP problem by solving many sub-problems iteratively.…”

Simultaneous production and distribution of industrial gas supply-chains

“…Also, refer to the recent paper of Coelho et al (2013) for the history of IRP and a review of different exact and heuristic approaches considered to solve wide variety of IRP problems. Taking into account customer demands, relevant IRP formulations include either deterministic (Dror et al, 1985;Chien et al, 1989;Jaillet et al 2002;Campbell et al 2004;Benoist et al, 2011) or stochastic (Dror and Ball, 1987;Çetinkaya and Lee, 2000;Kleywegt et al, 2002) approaches. The most simplified IRP is NP-hard as it contains the classical vehicle routing problem (VRP).…”