A variable neighborhood-based heuristic for the heterogeneous fleet vehicle routing problem

The heterogeneous fleet vehicle routing problem is investigated using some adaptations of the variable neighborhood search (VNS). The initial solution is obtained by Dijkstra’s algorithm based on a cost network constructed by the sweep algorithm and the 2-opt. Our VNS algorithm uses several neighborhoods which are adapted for this problem. In addition, a number of local search methods together with a diversification procedure are used. Two VNS variants, which differ in the order the diversification and Dijkstra’s algorithm are used, are implemented. Both variants appear to be competitive and produce new best results when tested on the data sets from the literature. We also constructed larger data sets for which benchmarking results are provided for future comparison.     

Underestimating the cost of a soft constraint is dangerous: revisiting the edit-distance based soft regular constraint

Many real-life problems are over-constrained, so that no solution satisfying all their constraints exists. Soft constraints, with costs denoting how much the constraints are violated, are used to solve these problems. We use the edit-distance based SoftRegular constraint as an example to show that a propagation algorithm that sometimes underestimates the cost may guide the search to incorrect (non-optimal) solutions to an over-constrained problem. To compute correctly the cost for the edit-distance based SoftRegular constraint, we present a quadratic-time propagation algorithm based on dynamic programming and a proof of its correctness. We also give an improved propagation algorithm using an idea of computing the edit distance between two strings, which may also be applied to other constraints with propagators based on dynamic programming. The asymptotic time complexity of our improved propagator is always at least as good as the one of our quadratic-time propagator, but significantly better when the edit distance is small. Our propagators achieve domain consistency on the problem variables and bounds consistency on the cost variable. Our method can also be adapted for the violation measure of the edit-distance based Regular constraint for constraint-based local search.     

Approximation in p-Norm of Univariate Concave Functions

We derive worst-case bounds, with respect to the L ^p norm, on the error achieved by algorithms aimed at approximating a concave function of a single variable, through the evaluation of the function and its subgradient at a fixed number of points to be determined. We prove that, for p larger than 1, adaptive algorithms outperform passive ones. Next, for the uniform norm, we propose an improvement of the Sandwich algorithm, based on a dynamic programming formulation of the problem.     

A Hamiltonian Krylov-Schur-type method based on the symplectic Lanczos process

We discuss a Krylov-Schur like restarting technique applied within the symplectic Lanczos algorithm for the Hamiltonian eigenvalue problem. This allows us to easily implement a purging and locking strategy in order to improve the convergence properties of the symplectic Lanczos algorithm. The Krylov-Schur-like restarting is based on the SR algorithm. Some ingredients of the latter need to be adapted to the structure of the symplectic Lanczos recursion. We demonstrate the efficiency of the new method for several Hamiltonian eigenproblems.     

Differential evolution for dynamic environments with unknown numbers of optima

This paper investigates optimization in dynamic environments where the numbers of optima are unknown or fluctuating. The authors present a novel algorithm, Dynamic Population Differential Evolution (DynPopDE), which is specifically designed for these problems. DynPopDE is a Differential Evolution based multi-population algorithm that dynamically spawns and removes populations as required. The new algorithm is evaluated on an extension of the Moving Peaks Benchmark. Comparisons with other state-of-the-art algorithms indicate that DynPopDE is an effective approach to use when the number of optima in a dynamic problem space is unknown or changing over time.     

A branch-and-cut algorithm for solving the Non-Preemptive Capacitated Swapping Problem

This paper models and solves a capacitated version of the Non-Preemptive Swapping Problem. This problem is defined on a complete digraph G=(V,A), at every vertex of which there may be one unit of supply of an item, one unit of demand, or both. The objective is to determine a minimum cost capacitated vehicle route for transporting the items in such a way that all demands are satisfied. The vehicle can carry more than one item at a time. Three mathematical programming formulations of the problem are provided. Several classes of valid inequalities are derived and incorporated within a branch-and-cut algorithm, and extensive computational experiments are performed on instances adapted from TSPLIB.     

Prepositioning can improve the performance of a dynamic stochastic on-demand public bus system

This paper presents the first application of prepositioning in the context of the dynamic stochastic on-demand bus routing problem (DODBRP). The DODBRP is a large-scale dial-a-ride problem that involves bus station assignment and aims to minimize the total user ride time (URT) by simultaneously assigning passengers to alternative stations and determining optimal bus routes.In the DODBRP, transportation requests are introduced dynamically, and buses are dispatched to stations with known requests. This paper investigates the concept of prepositioning, which involves sending buses not only to currently known requests but also to requests that are likely to appear in the future, based on a given probability.To solve this dynamic and stochastic ODBRP, the paper proposes a heuristic algorithm based on variable neighborhood search (VNS). The algorithm considers multiple scenarios to represent different realizations of the stochastic requests.Experimental results demonstrate the superiority of the prepositioning approach over the DODBRP across various levels of forecast accuracy, lengths of time bucket, and probabilities of realization. Furthermore, the paper shows that removing empty stations as a recourse action can further enhance solution quality. Additionally, in situations with low prediction accuracy, increasing the number of scenarios can lead to improved solutions. Finally, a combination of prepositioning, empty station removal, and the insertion of dynamic requests proves to be effective.Overall, the findings of this paper provide valuable insights into the application of prepositioning in the dynamic stochastic on-demand bus routing problem, highlighting its potential for addressing real-world transportation challenges.     

On an algorithm for dynamic reconstruction of the input

We consider the problem of dynamic reconstruction of the input in a system described by a vector differential equation and nonlinear in the state variable. We indicate an algorithm that is stable under information noises and computational errors and is aimed at infinite system operation time. The algorithm is based on the dynamic regularization method.     

Dynamic vehicle routing: Status and prospects

Although most real-world vehicle routing problems are dynamic, the traditional methodological arsenal for this class of problems has been based on adaptations of static algorithms. Still, some important new methodological approaches have recently emerged. In addition, computer-based technologies such as electronic data interchange (EDI), geographic information systems (GIS), global positioning systems (GPS), and intelligent vehicle-highway systems (IVHS) have significantly enhanced the possibilities for efficient dynamic routing and have opened interesting directions for new research. This paper examines the main issues in this rapidly growing area, and surveys recent results and other advances. The assessment of possible impact of new technologies and the distinction of dynamic problems vis-à-vis their static counterparts are given emphasis.     

Efficient Filtering and Tabu Search on a Consistent Neighbourhood for the Frequency Assignment Problem with Polarisation

This article presents a specific filtering algorithm for the Frequency Assignment Problem with Polarisation, which combines arc-consistency and path-inverse-consistency adapted to the specificities of the constraints. The effectiveness of this filtering algorithm enabled us to improve the Tabu Search on a Consistent Neighbourhood (CN-Tabu) using two different approaches. So, after a short recall of this general methodology and a presentation of its obtained results on the FAPP, we propose a behavioural study of the two approaches by comparing the results.     

Dynamic adaptation for parabolic equations

A dynamic adaptation method is presented that is based on the idea of using an arbitrary time-dependent system of coordinates that moves at a velocity determined by the unknown solution. Using some model problems as examples, the generation of grids that adapt to the solution is considered for parabolic equations. Among these problems are the nonlinear heat transfer problem concerning the formation of stationary and moving temperature fronts and the convection-diffusion problems described by the nonlinear Burgers and Buckley-Leverette equations. A detailed analysis of differential approximations and numerical results shows that the idea of using an arbitrary time-dependent system of coordinates for adapted grid generation in combination with the principle of quasi-stationarity makes the dynamic adaptation method universal, effective, and algorithmically simple. The universality is achieved due to the use of an arbitrary time-dependent system of coordinates that moves at a velocity determined by the unknown solution. This universal approach makes it possible to generate adapted grids for time-dependent problems of mathematical physics with various mathematical features. Among these features are large gradients, propagation of weak and strong discontinuities in nonlinear transport and heat transfer problems, and moving contact and free boundaries in fluid dynamics. The efficiency is determined by automatically fitting the velocity of the moving nodes to the dynamics of the solution. The close relationship between the adaptation mechanism and the structure of the parabolic equations allows one to automatically control the nodes’ motion so that their trajectories do not intersect. This mechanism can be applied to all parabolic equations in contrast to the hyperbolic equations, which do not include repulsive components. The simplicity of the algorithm is achieved due to the general approach to the adaptive grid generation, which is independent of the form and type of the differential equations.     

Run length not required: Optimal-mse dynamic batch means estimators for steady-state simulations

This paper addresses the estimation of the variance of the sample mean from steady-state simulations without requiring the knowledge of simulation run length a priori. Dynamic batch means is a new and useful approach to implementing the traditional batch means in limited memory without the knowledge of the simulation run length. However, existing dynamic batch means estimators do not allow one to control the value of batch size, which is the performance parameter of the batch means estimators. In this work, an algorithm is proposed based on two dynamic batch means estimators to dynamically estimate the optimal batch size as the simulation runs. The simulation results show that the proposed algorithm requires reasonable computation time and possesses good statistical properties such as small mean-squared-error (mse).     

A New Multiobjective Dynamic Routing Method for Multiservice Networks: Modelling and Performance

There are potential advantages in formulating the routing problems in modern multiservice networks as multiple objective problems. This paper presents a novel hierarchical bi-level multiobjective dynamic routing model for multiservice networks. It is based on a bi-objective shortest path algorithm, with dynamically adapted soft-constraints, to compute alternative paths for each node pair and on a heuristic to synchronously select alternative routing plans for the network in a dynamic alternative routing context. It is a routing method which periodically changes alternative paths as a function of periodic updates of certain QoS related parameters obtained from real-time measurements. The performance of the proposed routing method is compared with two reference dynamic routing methods namely RTNR and DAR by means of a discrete-event simulator.A previous short version of this work was presented at INOC’03 (International Network Optimisation Conference). Work partially supported by programme POSI of the III EC programme cosponsored by FEDER and national funds.     

L1 cheapest paths in “Fjord Scenery”

We show a simple proof of the existence of a path on the “border of water and rocks” based on combinatorial induction procedure and we present an algorithm for computing L₁ shortest path in “Fjord Scenery”. The proposed algorithm is a version of the Dijkstra technique adapted to a rectangle map with a square network. A few pre-processing modifications of the algorithm following from the combinatorial procedure are included. The validity of this approach is shown by numerical calculations for an example.     

A Recovering Beam Search algorithm for the one-machine dynamic total completion time scheduling problem

This paper deals with the one-machine dynamic total completion time scheduling problem. This problem is known to be NP-hard in the strong sense. A polynomial time heuristic algorithm is proposed which applies the recently introduced Recovering Beam Search (RBS) approach. The algorithm is based on a beam search procedure with unitary beam width and includes a recovering subroutine that allows to overcome wrong decisions taken at higher levels of the beam search tree. It is shown that the total number of considered nodes is bounded by n where n is the jobsize. The proposed algorithm is able to solve in very short CPU time problems with up to 500 jobs outperforming the best state of the art heuristics.     

Automatic deduction in (dynamic) geometry: Loci computation

A symbolic tool based on open source software that provides robust algebraic methods to handle automatic deduction tasks for a dynamic geometry construction is presented. The prototype has been developed as two different worksheets for the open source computer algebra system Sage, corresponding to two different ways of coding a geometric construction, namely with the open source dynamic geometry system GeoGebra or using the common file format for dynamic geometry developed by the Intergeo project. Locus computation algorithms based on Automatic Deduction techniques are recalled and presented as basic for an efficient treatment of advanced methods in dynamic geometry. Moreover, an algorithm to eliminate extraneous parts in symbolically computed loci is discussed. The algorithm, based on a recent work on the Gröbner cover of parametric systems, identifies degenerate components and extraneous adherence points in loci, both natural byproducts of general polynomial algebraic methods. Several examples are shown in detail.     

L p solutions of backward stochastic Volterra integral equations

This paper is devoted to the unique solvability of backward stochastic Volterra integral equations (BSVIEs, for short), in terms of both M-solution and the adapted solutions. We prove the existence and uniqueness of M-solutions of BSVIEs in L ^p (1 < p < 2), which extends the existing results on M-solutions. The unique solvability of adapted solutions of BSVIEs in L ^p (p > 1) is also considered, which also generalizes the results in the existing literature.     

Combining MFS and PGD methods to solve transient heat equation

We propose in this article a numerical algorithm based on the combination of the method of fundamental solutions (MFS) and the proper generalized decomposition technique (PGD) to solve time‐dependent heat equation. The MFS is considered as a truly meshless technique well adapted for a wide range of physical problems and the PGD approach can be considered as a reduction technique based on the separated representation of the variable functions. The proposed study relates to a separation between the spatial and temporal coordinates. To show the effectiveness of the proposed algorithm, several examples are presented and compared to the reference results.     

A hybrid metaheuristic approach to the university course timetabling problem

This paper describes the development of a novel metaheuristic that combines an electromagnetic-like mechanism (EM) and the great deluge algorithm (GD) for the University course timetabling problem. This well-known timetabling problem assigns lectures to specific numbers of timeslots and rooms maximizing the overall quality of the timetable while taking various constraints into account. EM is a population-based stochastic global optimization algorithm that is based on the theory of physics, simulating attraction and repulsion of sample points in moving toward optimality. GD is a local search procedure that allows worse solutions to be accepted based on some given upper boundary or ‘level’. In this paper, the dynamic force calculated from the attraction-repulsion mechanism is used as a decreasing rate to update the ‘level’ within the search process. The proposed method has been applied to a range of benchmark university course timetabling test problems from the literature. Moreover, the viability of the method has been tested by comparing its results with other reported results from the literature, demonstrating that the method is able to produce improved solutions to those currently published. We believe this is due to the combination of both approaches and the ability of the resultant algorithm to converge all solutions at every search process.     

Applications of resource assignment and scheduling with Petri Nets and heuristic search

This paper introduces a new Petri Net based approach for resource allocation and scheduling. The goals are (i) minimize the number of required resources given a set of jobs, (ii) find both an assignment for all jobs in the span of a predefined shift and (iii) the sequence in which such jobs are executed. The studied problem was inspired from a complex real life manufacturing shop as described in this document. The modeling of the processes and jobs is carried out with Petri Nets due to their capability of representing dynamic, concurrent discrete-event dynamic systems. The resource assignment starts with an initial feasible solution (initial number of resources) and then follows with a re-optimization process aimed to further reduce the resource requirements. The algorithm is based on a modified Heuristic Search method previously presented. The algorithm was tested first on a number of instances from the literature and then on the aforementioned system (a car seat cover manufacturer). The proposed approach shows not only good results in terms of performance but also shows the potential of Petri Nets for modeling and optimizing real-life systems. An implementation phase at the first stages of the process is underway at the time of writing.