Solving multiobjective vehicle routing problem with stochastic demand via evolutionary computation

This paper considers the routing of vehicles with limited capacity from a central depot to a set of geographically dispersed customers where actual demand is revealed only when the vehicle arrives at the customer. The solution to this vehicle routing problem with stochastic demand (VRPSD) involves the optimization of complete routing schedules with minimum travel distance, driver remuneration, and number of vehicles, subject to a number of constraints such as time windows and vehicle capacity. To solve such a multiobjective and multi-modal combinatorial optimization problem, this paper presents a multiobjective evolutionary algorithm that incorporates two VRPSD-specific heuristics for local exploitation and a route simulation method to evaluate the fitness of solutions. A new way of assessing the quality of solutions to the VRPSD on top of comparing their expected costs is also proposed. It is shown that the algorithm is capable of finding useful tradeoff solutions for the VRPSD and the solutions are robust to the stochastic nature of the problem. The developed algorithm is further validated on a few VRPSD instances adapted from Solomon’s vehicle routing problem with time windows (VRPTW) benchmark problems.     

Cyclic-order neighborhoods with application to the vehicle routing problem with stochastic demand

We examine neighborhood structures for heuristic search applicable to a general class of vehicle routing problems (VRPs). Our methodology utilizes a cyclic-order solution encoding, which maps a permutation of the customer set to a collection of many possible VRP solutions. We identify the best VRP solution in this collection via a polynomial-time algorithm from the literature. We design neighborhoods to search the space of cyclic orders. Utilizing a simulated annealing framework, we demonstrate the potential of cyclic-order neighborhoods to facilitate the discovery of high quality a priori solutions for the vehicle routing problem with stochastic demand (VRPSD). Without tailoring our solution procedure to this specific routing problem, we are able to match 16 of 19 known optimal VRPSD solutions. We also propose an updating procedure to evaluate the neighbors of a current solution and demonstrate its ability to reduce the computational expense of our approach.     

Using parallel & distributed computing for real-time solving of vehicle routing problems with stochastic demands

This paper focuses on the Vehicle Routing Problem with Stochastic Demands (VRPSD) and discusses how Parallel and Distributed Computing Systems can be employed to efficiently solve the VRPSD. Our approach deals with uncertainty in the customer demands by considering safety stocks, i.e. when designing the routes, part of the vehicle capacity is reserved to deal with potential emergency situations caused by unexpected demands. Thus, for a given VRPSD instance, our algorithm considers different levels of safety stocks. For each of these levels, a different scenario is defined. Then, the algorithm solves each scenario by integrating Monte Carlo simulation inside a heuristic-randomization process. This way, expected variable costs due to route failures can be naturally estimated even when customers’ demands follow a non-normal probability distribution. Use of parallelization strategies is then considered to run multiple instances of the algorithm in a concurrent way. The resulting concurrent solutions are then compared and the one with the minimum total costs is selected. Two numerical experiments allow analyzing the algorithm’s performance under different parallelization schemas.     

Vehicle routing problems with time windows and multiple service workers: a systematic comparison between ACO and GRASP

This paper systematically compares an ant colony optimization (ACO) and a greedy randomized adaptive search procedure (GRASP) metaheuristic. Both are used to solve the vehicle routing problem with time windows and multiple service workers. In order to keep the results comparable, the same route construction heuristic and local search procedures are used. It is shown that ACO clearly outperforms GRASP in the problem under study. Additionally, new globally best results for the used benchmark problems are presented.     

A hybrid Granular Tabu Search algorithm for the Multi-Depot Vehicle Routing Problem

In this paper, we propose a hybrid Granular Tabu Search algorithm to solve the Multi-Depot Vehicle Routing Problem (MDVRP). We are given on input a set of identical vehicles (each having a capacity and a maximum duration), a set of depots, and a set of customers with deterministic demands and service times. The problem consists of determining the routes to be performed to fulfill the demand of the customers by satisfying, for each route, the associated capacity and maximum duration constraints. The objective is to minimize the sum of the traveling costs related to the performed routes. The proposed algorithm is based on a heuristic framework previously introduced by the authors for the solution of the Capacitated Location Routing Problem (CLRP). The algorithm applies a hybrid Granular Tabu Search procedure, which considers different neighborhoods and diversification strategies, to improve the initial solution obtained by a hybrid procedure. Computational experiments on benchmark instances from the literature show that the proposed algorithm is able to produce, within short computing time, several best solutions obtained by the previously published methods and new best solutions.     

Speeding up continuous GRASP

Continuous GRASP (C-GRASP) is a stochastic local search metaheuristic for finding cost-efficient solutions to continuous global optimization problems subject to box constraints (Hirsch et al., 2007). Like a greedy randomized adaptive search procedure (GRASP), a C-GRASP is a multi-start procedure where a starting solution for local improvement is constructed in a greedy randomized fashion. In this paper, we describe several improvements that speed up the original C-GRASP and make it more robust. We compare the new C-GRASP with the original version as well as with other algorithms from the recent literature on a set of benchmark multimodal test functions whose global minima are known. Hart’s sequential stopping rule (1998) is implemented and C-GRASP is shown to converge on all test problems.     

A multi-space sampling heuristic for the vehicle routing problem with stochastic demands

The vehicle routing problem with stochastic demands consists in designing transportation routes of minimal expected cost to satisfy a set of customers with random demands of known probability distributions. This paper proposes a simple yet effective heuristic approach that uses randomized heuristics for the traveling salesman problem, a tour partitioning procedure, and a set partitioning formulation to sample the solution space and find high-quality solutions for the problem. Computational experiments on benchmark instances from the literature show that the proposed approach is competitive with the state-of-the-art algorithm for the problem in terms of both accuracy and efficiency. In experiments conducted on a set of 40 instances, the proposed approach unveiled four new best-known solutions (BKSs) and matched another 24. For the remaining 12 instances, the heuristic reported average gaps with respect to the BKS ranging from 0.69 to 0.15 % depending on its configuration.     

A Parallel Grasp for the Steiner Tree Problem in Graphs Using a Hybrid Local Search Strategy

In this paper, we present a parallel greedy randomized adaptive search procedure (GRASP) for the Steiner problem in graphs. GRASP is a two-phase metaheuristic. In the first phase, solutions are constructed using a greedy randomized procedure. Local search is applied in the second phase, leading to a local minimum with respect to a specified neighborhood. In the Steiner problem in graphs, feasible solutions can be characterized by their non-terminal nodes (Steiner nodes) or by their key-paths. According to this characterization, two GRASP procedures are described using different local search strategies. Both use an identical construction procedure. The first uses a node-based neighborhood for local search, while the second uses a path-based neighborhood. Computational results comparing the two procedures show that while the node-based variant produces better quality solutions, the path-based variant is about twice as fast. A hybrid GRASP procedure combining the two neighborhood search strategies is then proposed. Computational experiments with a parallel implementation of the hybrid procedure are reported, showing that the algorithm found optimal solutions for 45 out of 60 benchmark instances and was never off by more than 4% of the optimal solution value. The average speedup results observed for the test problems show that increasing the number of processors reduces elapsed times with increasing speedups. Moreover, the main contribution of the parallel algorithm concerns the fact that larger speedups of the same order of the number of processors are obtained exactly for the most difficult problems.     

A hybrid algorithm for the manufacturing cell formation problem

In this paper we compare different heuristic methods for the manufacturing cell formation problem considering part process sequence: a GRASP algorithm, a reactive GRASP algorithm and a hybrid algorithm which combines reactive GRASP and tabu search. All algorithms are tested with a set of instances from the literature. The results from the GRASP algorithm are compared to those of the reactive GRASP in order to evaluate the advantages of automatically adjusting the parameter value within the randomized greedy procedure. Also the reactive GRASP results are compared to those of the hybrid algorithm to evaluate the contribution to solution quality of replacing the local search phase of the GRASP algorithm with tabu search.     

A nonmonotone GRASP

A greedy randomized adaptive search procedure (GRASP) is an iterative multistart metaheuristic for difficult combinatorial optimization problems. Each GRASP iteration consists of two phases: a construction phase, in which a feasible solution is produced, and a local search phase, in which a local optimum in the neighborhood of the constructed solution is sought. Repeated applications of the construction procedure yields different starting solutions for the local search and the best overall solution is kept as the result. The GRASP local search applies iterative improvement until a locally optimal solution is found. During this phase, starting from the current solution an improving neighbor solution is accepted and considered as the new current solution. In this paper, we propose a variant of the GRASP framework that uses a new “nonmonotone” strategy to explore the neighborhood of the current solution. We formally state the convergence of the nonmonotone local search to a locally optimal solution and illustrate the effectiveness of the resulting Nonmonotone GRASP on three classical hard combinatorial optimization problems: the maximum cut problem (MAX-CUT), the weighted maximum satisfiability problem (MAX-SAT), and the quadratic assignment problem (QAP).     

Effective Heuristic Procedures for a Field Technician Scheduling Problem

This paper addresses a field technician scheduling problem faced by many service providers in telecommunication industry. The problem is to assign a set of jobs, at different locations with time windows, to a group of field technicians with different job skills. Such a problem can be viewed as a generalization of the well-known vehicle routing problem with time windows since technician skills need to be matched with job types. We designed and tested several heuristic procedures for solving the problem, namely a greedy heuristic, a local search algorithm, and a greedy randomized adaptive search procedure (GRASP). Our computational results indicate that GRASP is the most effective among them but requires more CPU time. However, the unique structure of GRASP allows us to exploit parallelism to achieve linear speed-up with respect to the number of machines used.     

Open vehicle routing problem with driver nodes and time deadlines

In this paper, we consider a variant of the open vehicle routing problem in which vehicles depart from the depot, visit a set of customers, and end their routes at special nodes called driver nodes. A driver node can be the home of the driver or a parking lot where the vehicle will stay overnight. The resulting problem is referred to as the open vehicle routing problem with driver nodes (OVRP-d). We consider three classes of OVRP-d: with no time constraints, with a maximum route duration, and with both a maximum route duration as well as time deadlines for visiting customers. For the solution of these problems, which are not addressed previously in the literature, we develop a new tabu search heuristic. Computational results on randomly generated instances indicate that the new heuristic exhibits a good performance both in terms of the solution quality and computation time.     

A reactive GRASP with path relinking for capacitated clustering

This paper presents a greedy randomized adaptive search procedure (GRASP) coupled with path relinking (PR) to solve the problem of clustering n nodes in a graph into p clusters. The objective is to maximize the sum of the edge weights within each cluster such that the sum of the corresponding node weights does not exceed a fixed capacity. In phase I, both a heaviest weight edge (HWE) algorithm and a constrained minimum cut algorithm are used to select seeds for initializing the p clusters. Feasible solutions are obtained with the help of a self-adjusting restricted candidate list that sequentially guides the assignment of the remaining nodes. At each major GRASP iteration, the list length is randomly set based on a probability density function that is updated dynamically to reflect the solution quality realized in past iterations. In phase II, three neighborhoods, each defined by common edge and node swaps, are explored to attain local optimality. The following exploration strategies are investigated: cyclic neighborhood search, variable neighborhood descent, and randomized variable neighborhood descent (RVND). The best solutions found are stored in an elite pool.     

A Tabu Search Approach for Delivering Pet Food and Flour in Switzerland

In this paper, we consider a real-life vehicle routeing problem that occurs in a major Swiss company producing pet food and flour. In contrast with usual hypothetical problems, a large variety of restrictions has to be considered. The main constraints are relative to the accessibility and the time windows at customers, the carrying capacities of vehicles, the total duration of routes and the drivers' breaks. To find good solutions to this problem, we propose two heuristic methods: a fast straightforward insertion procedure and a method based on tabu search techniques. Next, the produced solutions are compared with the routes actually covered by the company. Our outcomes indicate that the total distance travelled can be reduced significantly when such methods are used.     

Computing Approximate Solutions of the Maximum Covering Problem with GRASP

We consider the maximum covering problem, a combinatorial optimization problem that arises in many facility location problems. In this problem, a potential facility site covers a set of demand points. With each demand point, we associate a nonnegative weight. The task is to select a subset of p > 0 sites from the set of potential facility sites, such that the sum of weights of the covered demand points is maximized. We describe a greedy randomized adaptive search procedure (GRASP) for the maximum covering problem that finds good, though not necessarily optimum, placement configurations. We describe a well-known upper bound on the maximum coverage which can be computed by solving a linear program and show that on large instances, the GRASP can produce facility placements that are nearly optimal.     

Hybridized evolutionary local search algorithm for the team orienteering problem with time windows

The orienteering problem (OP) consists in finding an elementary path over a subset of vertices. Each vertex is associated with a profit that is collected on the visitor’s first visit. The objective is to maximize the collected profit with respect to a limit on the path’s length. The team orienteering problem (TOP) is an extension of the OP where a fixed number m of paths must be determined. This paper presents an effective hybrid metaheuristic to solve both the OP and the TOP with time windows. The method combines the greedy randomized adaptive search procedure (GRASP) with the evolutionary local search (ELS). ELS generates multiple distinct child solutions using a mutation mechanism. Each child solution is further improved by a local search procedure. GRASP provides multiple starting solutions to the ELS. The method is able to improve several best known results on available benchmark instances.     

An exact algorithm for a vehicle routing problem with time windows and multiple use of vehicles

The vehicle routing problem with multiple use of vehicles is a variant of the classical vehicle routing problem. It arises when each vehicle performs several routes during the workday due to strict time limits on route duration (e.g., when perishable goods are transported). The routes are defined over customers with a revenue, a demand and a time window. Given a fixed-size fleet of vehicles, it might not be possible to serve all customers. Thus, the customers must be chosen based on their associated revenue minus the traveling cost to reach them. We introduce a branch-and-price approach to address this problem where lower bounds are computed by solving the linear programming relaxation of a set packing formulation, using column generation. The pricing subproblems are elementary shortest path problems with resource constraints. Computational results are reported on euclidean problems derived from well-known benchmark instances for the vehicle routing problem with time windows.     

Solving the two-echelon location routing problem by a GRASP reinforced by a learning process and path relinking

The two-echelon location-routing problem (LRP-2E) arises from recent transportation applications like city logistics. In this problem, still seldom studied, first-level trips serve from a main depot a set of satellite depots, which must be located, while second-level trips visit customers from these satellites. After a literature review on the LRP-2E, we present four constructive heuristics and a hybrid metaheuristic: A greedy randomized adaptive search procedure (GRASP) complemented by a learning process (LP) and path relinking (PR). The GRASP and learning process involve three greedy randomized heuristics to generate trial solutions and two variable neighbourhood descent (VND) procedures to improve them. The optional path relinking adds a memory mechanism by combining intensification strategy and post-optimization. Numerical tests show that the GRASP with LP and PR outperforms the simple heuristics and an adaptation of a matheuristic initially published for a particular case, the capacitated location-routing problem (CLRP). Additional tests on the CLRP indicate that the best GRASP competes with the best metaheuristics published.     

The vehicle routing problem with flexible time windows and traveling times

We generalize the standard vehicle routing problem by allowing soft time window and soft traveling time constraints, where both constraints are treated as cost functions. With the proposed generalization, the problem becomes very general. In our algorithm, we use local search to determine the routes of vehicles. After fixing the route of each vehicle, we must determine the optimal start times of services at visited customers. We show that this subproblem is NP-hard when cost functions are general, but can be efficiently solved with dynamic programming when traveling time cost functions are convex even if time window cost functions are non-convex. We deal with the latter situation in the developed iterated local search algorithm. Finally we report computational results on benchmark instances, and confirm the benefits of the proposed generalization.     

Solving the pickup and delivery problem with three-dimensional loading constraints and reloading ban

In this paper, we extend the classical Pickup and Delivery Problem (PDP) to an integrated routing and three-dimensional loading problem, called PDP with three-dimensional loading constraints (3L-PDP). We are given a set of requests and a homogeneous fleet of vehicles. A set of routes of minimum total length has to be determined such that each request is transported from a loading site to the corresponding unloading site. In the 3L-PDP, each request is given as set of rectangular boxes and the vehicle capacity is replaced by a 3D loading space.This paper is the second one in a series of articles on 3L-PDP. As in the first paper we are dealing with constraints which guarantee that no reloading effort will occur. Here the focus is laid on the reloading ban, a packing constraint that ensures identical placements of same boxes in different packing plans. The reloading ban allows for better solutions in terms of travel distance than a routing constraint that was used in the first paper to preclude any reloading effort. To implement this packing constraint a new type of packing procedure is needed that is capable to generate a series of interrelated packing plans per route. This packing procedure, designed as tree search algorithm, and the corresponding concept of packing checks is the main contribution of the paper at hand. The packing procedure and a large neighborhood search procedure for routing form a hybrid algorithm for the 3L-PDP. Computational experiments were performed using 54 3L-PDP benchmark instances.