The demand for blood transfusion is considered a real world problem which is needed for various medical emergencies. The blood assignment problem was introduced to address this problem. The formulation of this problem stretches from managing critical blood shortage levels and blood unit expiration, to blood compatibility between donor and patients. Another contributing factor to the blood assignment problem, lies in the blood bank having to import additional blood units from external sources when supply cannot meet the demand. These challenges have serious consequences especially in the case where the demand for blood is very high. Taking these factors into consideration, this study implements a metaheuristic hybrid algorithm that combines symbiotic organisms search algorithm with the blood assignment policy in relation to the blood banks of South Africa. The aim of this study is to minimize blood product wastage with regards to expiration and importation, whilst maximizing product delivery to patients in need. In addition, this study also implements a unique way of generating randomized datasets based on social events relating to South Africa public holidays. The computational results indicate that the proposed hybrid algorithm performed well in minimizing blood importation, and experienced no form of expiration throughout the time period.
Ordering systems are a mechanism used to program the flow of production orders into the manufacturing system. The correct usage and parametrization of such systems have a significant impact on the performance of the production. One of the well-succeed ordering systems available in the literature is the Period-batch control (PBC), that allows one to group the orders into different production periods, and program it into the planning horizon. This paper assumes a manufacturing system controlled by PBC. On this system, this paper considered two performance indicators: a primary goal is to minimize the total tardiness and the second goal is to minimize the idleness of the production system. Two approaches are implemented to solve this problem: a mixed-integer programming model and eight algorithms based on the Iterated Greedy method. Beyond finding good results when comparing to the ones found by the mathematical model approach, this paper also performs the Tardiness × Production Capacity on each algorithm.
Stochastic Tunneling (STUN) is an optimization heuristic whose basic mechanism is based on reducing barriers for its search process between local optima via a non-linear transformation. Here, we hybridize STUN with the idea of Tabu Search (TS), namely, the avoidance of revisiting previously assessed solutions. This prevents STUN from inefficiently scan areas of the search space whose objective function values have already been "transformed away". We introduce the novel idea of using a probabilistic data structure (Bloom filters) to store a (quasi-)infinite tabu history. Empirical results for a combinatorial optimization problem show superior performance. An analysis of the tabu list statistics shows the importance of this hybridization idea.
We consider a framework for obtaining a sequence of converging primal and dual bounds based on mixed integer linear programming formulations on layered graphs. The proposed iterative algorithm avoids the typically rather large size of the full layered graph by approximating it incrementally. We focus in particular on this refinement step that extends the graph in each iteration. Novel path-based approaches are compared to existing variants from the literature. Experiments on two benchmark problems—the traveling salesman problem with time windows and the rooted distance-constrained minimum spanning tree problem—show the effectiveness of our new strategies. Moreover, we investigate the impact of a strong heuristic component within the algorithm, both for improving convergence speed and for improving the potential of an employed reduced cost fixing step.
A calibration laboratory studied in this research performs a thermal test that requires an analyst for setup and processing and an oven to perform such an essay. For convenience, it's possible to group some of the essays according to the oven capacity. In this scenario, this paper proposes a scheduling approach to minimize the total flowtime of the orders. This is a multiple resource scheduling problem, where a resource (operator) is used on two processes (oven setup and analysis). In contrast to the classical definition of multiple resource scheduling problems, the oven setup process requires the presence of the operator only for the startup of the process. To solve this problem, we derived: (i) a mixed-integer formulation; (ii) an Ant Colony Optimization (ACO) approach. On those developments, we also discuss some structural properties of this problem, that may lead to further advances in this field in the future. Our results show the ACO approach as a good alternative to the MIP, especially when solving instances with 30 service orders.
Construct, Merge, Solve & Adapt (CMSA) is a general hybrid metaheuristic for solving combinatorial optimization problems. At each iteration, CMSA (1) constructs feasible solutions to the tackled problem instance in a probabilistic way and (2) solves a reduced problem instance (if possible) to optimality. The construction of feasible solutions is hereby problem-specific, usually involving a fast greedy heuristic. The goal of this paper is to design a problem-agnostic CMSA variant whose exclusive input is an integer linear program (ILP). In order to reduce the complexity of this task, the current study is restricted to binary ILPs. In addition to a basic problem-agnostic CMSA variant, we also present an extended version that makes use of a constraint propagation engine for constructing solutions. The results show that our technique is able to match the upper bounds of the standalone application of CPLEX in the context of rather easy-to-solve instances, while it generally outperforms the standalone application of CPLEX in the context of hard instances. Moreover, the results indicate that the support of the constraint propagation engine is useful in the context of problems for which finding feasible solutions is rather difficult.
This study considers the problem of resource constrained project scheduling to maximise the net present value. A number of tasks must be scheduled within a fixed time horizon. Tasks may have precedences between them and they use a number of common resources when executing. For each resource, there is a limit, and the cumulative resource requirements of all tasks executing at the same time must not exceed the limits. To solve this problem, we develop a hybrid of Construct, Merge, Solve and Adapt (CMSA) and Ant Colony Optimisation (ACO). The methods are implemented in a parallel setting within a multi-core shared memory architecture. The results show that the proposed algorithm outperforms the previous state-of-the-art method, a hybrid of Lagrangian relaxation and ACO.
Traveling salesperson problem with hotel selection consists of determining a tour for the salesperson who needs to visit a predefined number of customers at different locations by taking into consideration that each working day is limited by time. If the time limit is accomplished, the salesperson must select a hotel from the set of available ones to spend the night. The aim is to minimize the number of necessary days to visit all customers spending the shortest possible travel time. We present an adaptive efficient heuristic based on the Iterated Local Search metaheuristic to solve available instances. The proposed heuristic is able to find good solutions for almost all instances and, in some cases, it is able to improve the quality of the best results found in literature, decreasing the number of trips necessary or time to travel along a tour. Moreover, the heuristic is fast enough to be applied to real problems that require fast responses.
In this article we present a comparison of the performance between a metaheuristic optimization method, Game of Patterns (GofP), so-called by the author, and the well-known genetic algorithms (GAs), through two implementations, namely: the GA of Scilab (SGA); and the GA of the R Project for Statistical Computing (RGA). For this purpose, we have selected a set of multimodal objective functions in the n-dimensional Euclidean space ℝ^n with a unique global minimum. For comparing both metaheuristic optimization approaches, a performance indicator of quality, denoted Q(p, n, s), was defined, which allows us to measure the quality of the obtained global optimal solution for each pth problem, in the n-dimensional space, when it is solved by each metaheuristic optimization method s∈{,,} . The indicator Q(p, n, s) then depends on: the number of evaluations of the pth optimization problem in the Euclidean space ℝ^n , which has required the s metaheuristic optimization method for identifying the global minimum; and the distance between the location of its respective unique global minimum and the location of the minimum that has been identified by the s metaheuristic optimization method. The paper also offers a brief explanation of the GofP method, which has been developed for solving unconstrained mixed integer problems in the n× m -dimensional Euclidean space ℝ^n×ℤ^m .
This work presents four hybrid methods based on the Self-adaptive Differential Evolution algorithm with fragment insertion applied to the protein structure prediction problem. The protein representation is the backbone torsion angles with side chain centroid coordinates. The fragment insertion is made by the Monte Carlo algorithm. The hybrid methods were compared with recent and compatible methods from the literature, where two proposed approaches achieved competitive results. The results have shown that using parameter control and fragment insertion greatly improves the results of the prediction when compared to fragment-less methods or without parameter control. Furthermore, an extra analysis was conducted using GDT-TS and TM-Score metrics to better understand the results obtained.
The Multi-Mode Resource-Constrained Project Scheduling Problem with energy saving (MRCPSP-energy) is a variant of the classical Resource-Constrained Project Scheduling Problem (RCPSP). In this variant, the execution of each job must take into account the job duration and the energy spent to execute that job, which are conflicting. The objective is to minimize both makespan and total energy consumption. This work proposes two local search methods to improve a large dataset of inputs. One of them is a restricted version of a Mixed-Integer Programming formulation and the other one is a heuristic local search called H. The computational experiments showed that the hybrid method with the H algorithm obtained better solutions and is competitive with the literature results.
Global energy demand has undergone a substantial increase in past decades because of the rapid increase of the global population and the energetic consumption of new production technologies. As a result, a change is necessary in the global energy generating matrix, in which the sources originate primarily from renewable energy sources. The main renewable energy source may be solar energy, and one of its applications is solar mobility. A world-class solar racing car exists that requires a rational use of velocity and energy to minimize the time spent in a race. A total of three search metaheuristics were tested to achieve an efficient velocity profile for this car in the Atacama 2018 Solar Race: Genetic Algorithm, Simulated Annealing and Iterated Local Search. The three methods provided similar results, with Simulated Annealing being the one that provided better solutions.
Office-space-allocation (OFA) problem is a category of a timetabling problem that involves the distribution of a set of limited entities to a set of resources subject to satisfying a set of given constraints. The constraints in OFA problem is of two types: hard and soft. The hard constraints are the one that must be satisfied for the solution to be feasible while the violation of soft constraints is allowed but it must be reduced as much as possible. The quality of the OFA solution is determined by the satisfaction of the soft constraints in a feasible solution. The complexity of the OFA problem motivated the researchers in the domain of AI and Operational research to develop numerous metaheuristic-based techniques. Among recently introduced local search-based metaheuristic techniques that have been successfully utilized to solve complex optimization problem is the Late Acceptance Hill Climbing (LAHC) algorithm. This paper presents an adaptation of LAHC algorithm to tackle the OFA problem in which three neighbourhood structures are embedded with the operators of the LAHC algorithm in order to explore the solution space of the OFA efficiently. The benchmark instances proposed by the University of Nottingham and University of Wolverhampton datasets are employed in the evaluation of the proposed algorithm. The LAHC algorithm is able to produced one new result, two best results and competitive results when compared with the state-of-the-art methods.
In this paper we present a new population based metaheuristic called the fixed set search (FSS). The proposed approach represents a method of adding a learning mechanism to the greedy randomized adaptive search procedure (GRASP). The basic concept of FSS is to avoid focusing on specific high quality solutions but on parts or elements that such solutions have. This is done through fixing a set of elements that exist in such solutions and dedicating computational effort to finding near optimal solutions for the underlying subproblem. The simplicity of implementing the proposed method is illustrated on the traveling salesman problem. Our computational experiments show that the FSS manages to find significantly better solutions than the GRASP it is based on, the dynamic convexized method and the ant colony optimization combined with a local search.
In this paper, two heuristic algorithms are proposed to solve the direct aperture optimisation problem (DAO) in radiation therapy for cancer treatment. In the DAO problem, the goal is to find a set of deliverable aperture shapes and intensities so we can irradiate the tumor according to a medical prescription without producing any harm to the surrounding healthy tissues. Unlike the traditional two-step approach used in intensity modulated radiation therapy (IMRT) where the intensities are computed and then the apertures shapes are determined by solving a sequencing problem, in the DAO problem, constraints associated to the number of deliverable aperture shapes as well as physical constraints are taken into account during the intensities optimisation process. Thus, we do not longer need any leaves sequencing procedure after solving the DAO problem. We try our heuristic algorithms on a prostate case and compare the obtained treatment plan to the one obtained using the traditional two-step approach. Results show that our algorithms are able to find treatment plans that are very competitive when considering the number of deliverable aperture shapes.
There is a strong interplay between network reliability and connectivity theory. In fact, previous studies show that the graphs with maximum reliability, called uniformly most-reliable graphs, must have the highest connectivity. In this paper, we revisit the underlying theory in order to build uniformly most-reliable cubic graphs. The computational complexity of the problem promotes the development of heuristics. The contributions of this paper are three-fold. In a first stage, we propose an ideal Variable Neighborhood Descent (VND) which returns the graph with maximum reliability. This VND works in exponential time. In a second stage, we propose a hybrid GRASP/VND approach that trades quality for computational effort. A construction phase enriched with a Restricted Candidate List (RCL) offers diversification. Our local search phase includes a factor-2 algorithm for an Integer Linear Programming (ILP) model. As a product of our research, we recovered previous optimal graphs from the related literature in the field. Additionally, we offer new candidates of uniformly most-reliable graphs with maximum connectivity and maximum number of spanning trees.
The automated transit networks (ATN) is a new and sophisticated concept which has the possibility to solve problems related to transit in urban areas. In ATN, driverless vehicles run on exclusive guideways in order to provide on-demand transportation service. In this paper, we focus on the strategic level of decision related to ATN. We deal with the problem of determining the best size of fleet of ATN vehicles while satisfying a set of transportation demands. A hybrid heuristic approach is developed while taking into account the objective of finding good quality solutions in a short computational time. Computational results performed in this study demonstrate the efficiency of our approach.
Seaside operations at container ports often suffer from uncertainty due to events such as the variation in arrival and/or processing time of vessels, weather conditions and others. Finding a robust plan which can accommodate this uncertainty is therefore desirable to port operators. This paper suggests ways to generate robust berth allocation plans in container terminals. The problem is first formulated as a mixed-integer programming model whose main objective is to minimize the total tardiness of vessel departure time. It is then solved exactly and approximately. Experimental results show that only small instances of the proposed model can be solved exactly. To handle large instances in reasonable times, the Genetic Algorithm (GA) is used. However, it does not guarantee optimality and often the approximate solutions returned are of low quality. A hybrid meta-heuristic which combines Branch-and-Cut (B&C) as implemented in CPLEX, with the GA as we implement it here, is therefore suggested. This hybrid method retains the accuracy of Branch-and-Cut and the efficiency of GA. Numerical results obtained with the three approaches on a representative set of instances of the problem are reported.
This paper deals with novel approach for hybridization of two scientific techniques: the evolutionary computational techniques and deterministic chaos. The Particle Swarm Optimization algorithm is enhanced with two pseudo-random number generators based on chaotic systems. The chaotic pseudo-random number generators (CPRNGs) are used to guide the particles movement through multiplying the accelerating constants. Different CPRNGs are used simultaneously in order to improve the performance of the algorithm. The IEEE CEC’13 benchmark suite is used to test the performance of the proposed method.
Hexapod robots are six-legged robotic systems, which have been widely investigated in the literature for various applications including exploration, rescue, and surveillance. Designing hexapod robots requires to carefully considering a number of different aspects. One of the aspects that require careful design attention is the planning of leg trajectories. In particular, given the high demand for fast motion and high-energy autonomy it is important to identify proper leg operation paths that can minimize energy consumption while maximizing the velocity of the movements. In this frame, this paper presents a preliminary study on the application of a hybrid multi-objective optimization approach for the computer-aided optimal design of a legged robot. To assess the methodology, a kinematic and dynamic model of a leg of a hexapod robot is proposed as referring to the main design parameters of a leg. Optimal criteria have been identified for minimizing the energy consumption and efficiency as well as maximizing the walking speed and the size of obstacles that a leg can overtake. We evaluate the performance of the hybrid multi-objective evolutionary approach to explore the design space and provide a designer with an optimal setting of the parameters. Our simulations demonstrate the effectiveness of the hybrid approach by obtaining improved Pareto sets of trade-off solutions as compared with a standard evolutionary algorithm. Computational costs show an acceptable increase for an off-line path planner.
