Multi-commodity location-routing: Flow intercepting formulation and branch-and-cut algorithm

We propose a novel hub location model that jointly eliminates some of the traditional assumptions on the structure of the network and on the discount as a result of economies of scale in an effort to better reflect real-world logistics and transportation systems. Our model extends the hub literature in various facets: instead of connecting nonhub nodes directly to hub nodes, we consider routes with stopovers; instead of connecting pairs of hubs directly, we design routes that can visit several hub nodes; rather than dimensioning pairwise connections, we dimension routes of vehicles; and rather than working with a homogeneous fleet, we use intermodal transportation. Decisions pertinent to strategic and tactical hub location and transportation network design are concurrently made through the proposed optimization scheme. An effective branch-and-cut algorithm is developed to solve realistically sized problem instances and to provide managerial insights.

Download Full-text

Metodología para creación de rutas alimentadoras en sistemas de transporte masivo

Revista Facultad de Ingeniería ◽

10.19053/01211129.v26.n45.2017.6052 ◽

2017 ◽

Vol 26 (45) ◽

Author(s):

Daniela Ospina-Toro ◽

Eliana Mirledy Toro-Ocampo ◽

Ramón Alfonso Gallego-Rendón

Keyword(s):

Programming Model ◽

Methodological Approach ◽

Exact Algorithm ◽

Set Partitioning ◽

Branch And Cut ◽

Mixed Integer ◽

Location Routing Problem ◽

Transit System ◽

Routing Problem ◽

Location Routing

This paper proposes a methodology to identify feeder routes for areas disconnected to the Mass Transit System (MTS), in order to propose an alternative solution to the deficit in the number of passengers carried. The proposed methodology consists of two steps: (1) structuring scenarios for areas not connected to the transport system and (2) combining heuristic and exact techniques to solve the feeding routes problem considering in the restrictions the path length and passengers vehicle capacity. To model the problem, a comparison with the Location Routing problem is established, which is usually applied to freight transport problems. The methodology proposed is a math-heuristic combining the Lin-Kernighan-Helsgaun algorithm (LKH) and the Clark and Wright’s Savings heuristic with the Branch-and-Cut exact algorithm, which is applied into a Mixed Integer Linear Programming model (MILP), also known as a Set Partitioning model (SP) for LRP. This methodological approach is validated with real instances considering locations in Pereira (Megabús), where some areas disconnected to the Central-Occidental Metropolitan Area System (AMCO) of Pereira, located in Colombia's Coffee Axis are considered.

Download Full-text

Fuzzy multi-objective location and routing problem

Journal of Intelligent & Fuzzy Systems ◽

10.3233/jifs-191654 ◽

2020 ◽

Vol 39 (3) ◽

pp. 3259-3273

Author(s):

Nasser Shahsavari-Pour ◽

Najmeh Bahram-Pour ◽

Mojde Kazemi

Keyword(s):

Firefly Algorithm ◽

High Reliability ◽

Research Area ◽

Nsga Ii ◽

Location Routing Problem ◽

Routing Problem ◽

Location Allocation ◽

Multi Objective ◽

Location Routing ◽

The Cost

The location-routing problem is a research area that simultaneously solves location-allocation and vehicle routing issues. It is critical to delivering emergency goods to customers with high reliability. In this paper, reliability in location and routing problems was considered as the probability of failure in depots, vehicles, and routs. The problem has two objectives, minimizing the cost and maximizing the reliability, the latter expressed by minimizing the expected cost of failure. First, a mathematical model of the problem was presented and due to its NP-hard nature, it was solved by a meta-heuristic approach using a NSGA-II algorithm and a discrete multi-objective firefly algorithm. The efficiency of these algorithms was studied through a complete set of examples and it was found that the multi-objective discrete firefly algorithm has a better Diversification Metric (DM) index; the Mean Ideal Distance (MID) and Spacing Metric (SM) indexes are only suitable for small to medium problems, losing their effectiveness for big problems.

Download Full-text