Well-tuned algorithms for the Team Orienteering Problem with Time Windows

The Team Orienteering Problem with Time Windows (TOPTW) is the extension of the Orienteering Problem (OP) where each node is limited by a predefined time window during which the service has to start. The objective of the TOPTW is to maximize the total collected score by visiting a set of nodes with a limited number of paths. We propose two algorithms, Iterated Local Search and a hybridization of Simulated Annealing and Iterated Local Search (SAILS), to solve the TOPTW. As indicated in multiple research works on algorithms for the OP and its variants, determining appropriate parameter values in a statistical way remains a challenge. We apply Design of Experiments, namely factorial experimental design, to screen and rank all the parameters thereby allowing us to focus on the parameter search space of the important parameters. The proposed algorithms are tested on benchmark TOPTW instances. We demonstrate that well-tuned ILS and SAILS lead to improvements in terms of the quality of the solutions. More precisely, we are able to improve 50 best known solution values on the available benchmark instances.     

The Orienteering Problem with Time Windows

The orienteering problem with time windows, denoted by OPTW, belongs to a class of routeing and scheduling problems that arise in physical distribution. It may be modelled as a problem on a graph. It considers a set of nodes (customers), each with an associated profit and service duration (time window), and a set of arcs, each with an associated travel time. The objective of the problem is to construct an acyclic path beginning at a specified origin and ending at a specified destination that maximizes the total profit while observing time window constraints on all nodes and not exceeding a designated time limit. The problem is classified as NP-hard and, thus, an exact algorithm that executes in reasonable computational time is unlikely to exist. Since the problem is highly-constrained, we were able to develop a heuristic (referred to as the ‘tree’ heuristic) based upon an exhaustive search of the feasible solution space. The tree heuristic systematically generates a list of feasible paths and then selects the most profitable path from the list. In comparison with an insertion heuristic, the tree heuristic was found to produce improved values of total profit for heavily-constrained, modest-sized problems with reasonable computational effort.     

An iterative three-component heuristic for the team orienteering problem with time windows

This paper studies the team orienteering problem with time windows, the aim of which is to maximize the total profit collected by visiting a set of customers with a limited number of vehicles. Each customer has a profit, a service time and a time window. A service provided to any customer must begin in his or her time window. We propose an iterative framework incorporating three components to solve this problem. The first two components are a local search procedure and a simulated annealing procedure. They explore the solution space and discover a set of routes. The third component recombines the routes to identify high quality solutions. Our computational results indicate that this heuristic outperforms the existing approaches in the literature in average performance by at least 0.41%. In addition, 35 new best solutions are found.     

A synthesis of assignment and heuristic solutions for vehicle routing with time windows

This paper proposes a three-stage method for the vehicle-routing problem with time window constraints (VRPTW). Using the Hungarian method the optimal customer matching for an assignment approximation of the VRPTW, which is a travel time-based relaxation that partially respects the time windows, is obtained. The assignment matching is transformed into feasible routes of the VRPTW via a simple decoupling heuristic. The best of these routes, in terms of travelling and vehicle waiting times, form part of the final solution, which is completed by the routes provided by heuristic methods applied to the remainder of the customers. The proposed approach is tested on a set of standard literature problems, and improves the results of the heuristic methods with respect to total travel time. Furthermore, it provides useful insights into the effect of employing optimal travel time solutions resulting from the assignment relaxation to derive partial route sets of the VRPTW.     

Heuristics for the multi-depot petrol station replenishment problem with time windows

In the multi-depot petrol station replenishment problem with time windows (MPSRPTW), the delivery of petroleum products stored in a number of different petroleum depots to a set of petrol distribution stations has to be optimized. Each depot has its own fleet of heterogeneous and compartmented tank trucks. Stations specify their demand by indicating the minimum and maximum quantities to be delivered for each ordered product and require the delivery within a predetermined time window. Several inter-related decisions must be made simultaneously in order to solve the problem. For this problem, the set of feasible routes to deliver all the demands, the departure depot for each route, the quantities of each product to be delivered, the assignment of these routes to trucks, the time schedule for each trip, and the loading of the ordered products to different tanks of the trucks used need to be determined. In this paper, we propose a mathematical model that selects, among a set of feasible trips, the subset that allows the delivery of all the demands while maximizing the overall daily net revenue. If this model is provided with all possible feasible trips, it determines the optimal solution for the corresponding MPSRPTW. However, since the number of such trips is often huge, we developed a procedure to generate a restricted set of promising feasible trips. Using this restricted set, the model produces a good but not necessarily optimal solution. Thus the proposed solution process can be seen as a heuristic. We report the results of the extensive numerical tests carried out to assess the performance of the proposed heuristic. In addition, we show that, for the special case of only one depot, the proposed heuristic outperforms a previously published solution method.     

An efficient heuristic algorithm for the capacitated p-\! median problem

The capacitated $p$ -median problem (CPMP) is one of the well-known facility-location problems. The objective of the problem is to minimize total cost of locating a set of capacitated service points and allocating a set of demand points to the located service points, while the total allocated demand for each service point is not be greater than its capacity limit. This paper presents an efficient heuristic algorithm based on the local branching and relaxation induced neighborhood search methods for the CPMP. The proposed algorithm is a heuristic technique that utilizes a general mixed integer programming solver to explore neighborhoods. The parameters of the proposed algorithm are tuned by design of experiments. The proposed method is tested on a large set of benchmark instances. The results show that the method outperforms the best method found in the literature.     

A mathematical model for the Multi-Levels Product Allocation Problem in a warehouse with compatibility constraints

The aim of this work is to address the products allocation problem in a multi-layers warehouse with compatibility constraints among the classes. The problem under study represents one of the most relevant topic in Logistics. The goal is to reduce, as much as possible, the delivery times; the inventories; the total logistic costs and to guarantee, at the same time, higher service levels (i.e., high customers satisfaction degree). In this work, a linear model to mathematically represent the problem is developed and its performance is evaluated on a set of instances, representing realistic situations. A sensitivity analysis is also carried out by considering the most relevant parameters of the model. Finally, an Iterated Local Search based heuristic is defined in order to solve large scale scenarios in a reasonable amount of time. Numerical results show that the proposed heuristic is able to find good quality solutions with a computational effort lower than that required to solve the proposed mathematical model.     

A time-based pheromone approach for the ant system

The ant system (AS) is a metaheuristic approach originally developed for solving the traveling salesman problem. AS has been successfully applied to various hard combinatorial optimization problems and different variants have been proposed in the literature. In this paper, we introduce a time-based pheromone approach for AS (TbAS). Due to this nature TbAS is applicable to routing problems involving time-windows. The novelty in TbAS is the multi-layer pheromone network structure which implicitly utilizes the service time information associated with the customers as a heuristic information. To investigate the performance of TbAS, we use the well-known vehicle routing problem with time-windows as our testbed and we conduct an extensive computational study using the Solomon (Algorithms for the vehicle routing and scheduling problems with time window constraints 35:254?C265, 1987) instances. Our results reveal that the proposed time-based pheromone approach is effective in obtaining good quality solutions.     

Vehicle routing with multiple deliverymen: Modeling and heuristic approaches for the VRPTW

In real life distribution of goods, relatively long service times may make it difficult to serve all requests during regular working hours. These difficulties are even greater if the beginning of the service in each demand site must occur within a time window and violations of routing time restrictions are particularly undesirable. We address this situation by considering a variant of the vehicle routing problem with time windows for which, besides routing and scheduling decisions, a number of extra deliverymen can be assigned to each route in order to reduce service times. This problem appears, for example, in the distribution of beverage and tobacco in highly dense Brazilian urban areas. We present a mathematical programming formulation for the problem, as well as a tabu search and an ant colony optimization heuristics for obtaining minimum cost routes. The performance of the model and the heuristic approaches are evaluated using instances generated from a set of classic examples from the literature.     

A heuristic method for dispatching repair men

A company has to provide service to its customers. A service consists of a visit to the customer plus the spending of some given time at the scene. The future customer demand is not known but the probability distribution for the demand may be known. When a customer call comes in, the company must immediately specify a time window within which the start of service will be provided. The problem is for a fixed service level to determine an optimal strategy of route design and time window setting so that the total distance travelled is minimized over the time horizon given. A heuristic method BARTOC (Booking Algorithm for Routing and Timing Of Customers) to solve the problem mentioned above is suggested. BARTOC is based on a cluster-first route-second approach. Some computational results are presented. The results indicate that BARTOC produces high quality solutions.Peter Matthiesen, Inc.Dano Chemo, Inc.     

Simple Heuristics for the Vehicle Routeing Problem with Soft Time Windows

We describe three simple heuristics for the vehicle routeing problem with customer time windows that can be violated by paying appropriate penalties. The customer demands are known, and a homogeneous fleet of vehicles stationed at a single depot is considered. The penalty payable to a customer is assumed to be a linear function of the amount of time window violation. Upper limits are imposed on both the penalty payable and the waiting time allowed at any customer. At each customer in a route, the PC-based heuristics simultaneously determine the actual time to begin service, and the next customer to serve. To achieve this, each heuristic uses different measures to compare the cost of waiting and penalty payable, with the benefit obtained by leaving immediately for the next customer. Computational results on a set of benchmark problems show that the procedure is efficient and enables significant reduction in the number of vehicles required and/or the total route distances while controlling both customer penalties and waiting times.     

Scheduling multi-stage parallel-processor services to minimize average response time

The problem of scheduling on a multi-stage parallel-processor architecture in computer centres is addressed with the objective of minimizing average completion time of a set of requests. The problem is modelled as a flexible flowshop problem for which few heuristics exist in the flowshop scheduling literature. A new three-phase heuristic is proposed in this paper. An extensive computational experiment has been conducted to compare the performance of the existing heuristics and the proposed heuristic. The results indicate that the proposed heuristic significantly outperforms the existing ones. More specifically, the overall average error of the best existing heuristic is about five times that of the proposed heuristic while the overall average CPU time of the proposed heuristic is about half of the best existing one. More importantly, as the number of requests increases, the CPU time of the proposed heuristic decreases considerably (compared to the best existing heuristic) while the ratio of the error (of the best existing to the proposed heuristic) of about five times remains almost the same.     

A GRASP heuristic for the manufacturing cell formation problem

In this work, we address the Manufacturing Cell Formation Problem (MCFP). Cellular Manufacturing is a production strategy that has emerged to reduce materials handling and set up times in order to reduce lead times in production systems and to improve customer??s service levels while reducing costs. We propose a GRASP heuristic to obtain lower bounds for the optimal solution of the problem. To evaluate the performance of the proposed method, we test the heuristic with different instances from the literature and compare the results obtained with those provided by other heuristic methods from the literature. According to the obtained results, the proposed GRASP procedure provides good quality lower bounds with reasonable computational effort.     

Routing relatively few customers per route

This paper presents a modification of the well-known Solomon (1987) sequential insertion heuristic I1 for the Vehicle Routing Problem with Time Windows (VRPTW). VRPTW involves servicing customers within a pre-specified service time window by homogeneously capacitated vehicles from a single depot. By using two new measures for the additional time needed to insert a customer in the route, significantly better solutions are obtained for relatively short routes compared to the original Solomon (1987) sequential insertion heuristic I1. Because high-quality initial heuristics often allow local searches and metaheuristics to achieve better solutions more quickly, the new approach is likely to help generating better solutions to practical routing problems.     

Managing pharmaceuticals delivery service using a hybrid particle swarm intelligence approach

Medicines or drugs have unique characteristics of short life cycle, small size, light weight, restrictive distribution time and the need of temperature and humidity control (selected items only). Thus, logistics companies often use different types of vehicles with different carrying capacities, and considering fixed and variable costs in service delivery, which make the vehicle assignment and route optimization more complicated. In this study, we formulate the problem to a multi-type vehicle assignment and mixed integer programming route optimization model with fixed fleet size under the constraints of distribution time and carrying capacity. Given non-deterministic polynomial hard and optimal algorithm can only be used to solve small-size problem, a hybrid particle swarm intelligence (PSI) heuristic approach, which adopts the crossover and mutation operators from genetic algorithm and 2-opt local search strategy, is proposed to solve the problem. We also adapt a principle based on cost network and Dijkstra’s algorithm for vehicle scheduling to balance the distribution time limit and the high loading rate. We verify the relative performance of the proposed method against several known optimal or heuristic solutions using a standard data set for heterogeneous fleet vehicle routing problem. Additionally, we compare the relative performance of our proposed Hybrid PSI algorithm with two intelligent-based algorithms, Hybrid Population Heuristic algorithm and Improved Genetic Algorithm, using a real-world data set to illustrate the practical and validity of the model and algorithm.

     

Fine-tuning a parametric Clarke and Wright heuristic by means of EAGH (empirically adjusted greedy heuristics)

Alt?nel and Öncan (2005) (A new enhancement of the Clarke and Wrightsavings heuristic for the capacitated vehicle routing problem) proposed aparametric Clarke and Wright heuristic to solve the capacitated vehicle routingproblem (CVRP). The performance of this parametric heuristic is sensitive tofine-tuning. Antinel and Öncan used an enumerative parameter-settingapproach and improved on the results obtained with the original Clarke andWright heuristic, but their approach requires much more computation time tosolve an instance. Battarra et al (2008) (Tuning a parametricClarke–Wright heuristic through a genetic algorithm) proposed a geneticalgorithm to set the parameter values. They succeeded in reducing the timeneeded to solve an instance, but the quality of the solution was slightly worse.In this paper, we propose to use the EAGH (empirically adjusted greedyheuristics) procedure to set the parameter values. A computational experimentshows the efficiency of EAGH; in an even shorter time, we improve on the bestresults obtained with any parametric Clarke and Wright heuristic method proposedin the literature.     

Integrated service network design for a cross-docking supply chain network

This paper considers an integrated service network design problem for a given set of freight demands that is concerned with integration of locating cross-docking (CD) centers and allocating vehicles for the associated direct (transportation) services from origin node to a CD center or from a CD center to the destination node. For the vehicle allocation, direct services (sub-routes) should be determined for the given freight demands, and then the vehicle allocation has to be made in consideration of routing for the associated direct service fulfillment subject to vehicle capacity and service time restriction. The problem is modeled as a path-based formulation for which a tabu-search-based solution algorithm is proposed. To guarantee the performance of the proposed solution algorithm, strong valid inequalities are derived based on the polyhedral characteristics of the problem domain and an efficient separation heuristic is derived for identifying any violated valid inequalities. Computational experiments are performed to test the performance of the proposed solution algorithm and also that of a valid-inequality separation algorithm, which finds that the solution algorithm works quite well and the separation algorithm provides strengthened lower bounds. Its immediate application may be made to strategic (integrated) service network designs and to tactical service network planning for the CD network.     

需要安装时间的平行多功能机排序问题的启发式算法

考虑需要安装时间的平行多功能机排序问题。在该模型中,每个工件对应机器集合的一个子集,其只能在这个子集中的任一台机器上加工,称这个子集为该工件的加工集合;工件分组,同组工件具有相同的加工时间和加工集合,不同组中的工件在同一台机器上连续加工需要安装时间,目标函数为极小化最大完工时间。对该问题NP-难的一般情况设计启发式算法:首先按照特定的规则将所有工件组都整组地安排到各台机器上,然后通过在各机器间转移工件不断改进当前最大完工时间。通过与下界的比较检验算法的性能,大量的计算实验表明,算法是实用而有效的。     

Vehicle routing with cross-docking

Over the past decade, cross-docking has emerged as an important material handling technology in transportation. A variation of the well-known Vehicle Routing Problem (VRP), the VRP with Cross-Docking (VRPCD) arises in a number of logistics planning contexts. This paper addresses the VRPCD, where a set of homogeneous vehicles are used to transport orders from the suppliers to the corresponding customers via a cross-dock. The orders can be consolidated at the cross-dock but cannot be stored for very long because the cross-dock does not have long-term inventory-holding capabilities. The objective of the VRPCD is to minimize the total travel time while respecting time window constraints at the nodes and a time horizon for the whole transportation operation. In this paper, a mixed integer programming formulation for the VRPCD is proposed. A tabu search heuristic is embedded within an adaptive memory procedure to solve the problem. The proposed algorithm is implemented and tested on data sets provided by the Danish consultancy Transvision, and involving up to 200 pairs of nodes. Experimental results show that this algorithm can produce high-quality solutions (less than 5% away from optimal solution values) within very short computational time.     

The Team Orienteering Problem with Time Windows: An LP-based Granular Variable Neighborhood Search

The Team Orienteering Problem (TOP) is a known NP-hard problem that typically arises in vehicle routing and production scheduling contexts. In this paper we introduce a new solution method to solve the TOP with hard Time Window constraints (TOPTW). We propose a Variable Neighborhood Search (VNS) procedure based on the idea of exploring, most of the time, granular instead of complete neighborhoods in order to improve the algorithm’s efficiency without loosing effectiveness. The method provides a general way to deal with granularity for those routing problems based on profits and complicated by time constraints. Extensive computational results are reported on standard benchmark instances. Performance of the proposed algorithm is compared to optimal solution values, when available, or to best known solution values obtained by state-of-the-art algorithms. The method comes out to be, on average, quite effective allowing to improve the best know values for 25 test instances.