Ant Colony Optimization and Swarm Intelligence

In this paper we report an empirical comparison of some of the most influential Particle Swarm Optimization (PSO) algorithms based on run-length distributions (RLDs). The advantage of our approach over the usual report pattern (average iterations to reach a predefined goal, success rates, and standard deviations) found in the current PSO literature is that it is possible to evaluate the performance of an algorithm on different application scenarios at the same time. The RLDs reported in this paper show some of the strengths and weaknesses of the studied algorithms and suggest ways of improving their performance.

A theoretical framework and model is presented to study the self-organized behavior of probabilistic routing protocols for computer networks. Such soft routing protocols have attracted attention for delivering packets reliably, robustly, and efficiently. The framework supports several features necessary for emergent routing behavior, including feedback loops and indirect communication between peers. Efficient global operating parameters can be estimated without resorting to expensive monte-carlo simulation of the whole system. Key model parameters are routing sensitivity and routing threshold, or noise, which control the “randomness” of packet routes between source and destination, and a metric estimator. Global network characteristics are estimated, including steady state routing probabilities, average path length, and path robustness.

The framework is based on a markov chain analysis. Individual network nodes are represented as states. Standard techniques are used to find primary statistics of the steady state global routing pattern, given a set of link costs. The use of packets to collect information about, or “sample,” the network for new path information is also reviewed. How the network sample rate influences performance is investigated.

In this paper we propose a Stochastic User Equilibrium (SUE) algorithm that can be adopted as a model, known as a simulation model, that imitates the behaviour of transportation systems. Indeed, analyses of real dimension networks need simulation algorithms that allow network conditions and performances to be rapidly determined. Hence, we developed an MSA (

Method of Successive Averages

) algorithm based on the Ant Colony Optimisation paradigm that allows transportation systems to be simulated in less time but with the same accuracy as traditional MSA algorithms. Finally, by means of Blum’s theorem, we stated theoretically the convergence of the proposed ACO-based algorithm.

Mobile ad hoc networks are a class of highly dynamic networks. In previous work, we developed a new routing algorithm, called AntHocNet, for these challenging network environments. AntHocNet has been designed after the Ant Colony Optimization (ACO) framework, and its general architecture shares strong similarities with the architectures of typical ACO implementations for network routing. On the other hand, AntHocNet also contains several elements which are new to ACO routing implementations, such as the combination of ant-based path sampling with a lightweight information bootstrapping process, the use of both reactive and proactive components, and the use of composite pheromone metrics. In this paper we discuss all these elements, pointing out their general usefulness to face the multiple challenges of mobile ad hoc networks, and perform an evaluation of their working and effect on performance through extensive simulation studies.

Wireless Sensor Networks are characterized by having specific requirements such as limited energy availability, low memory and reduced processing power. On the other hand, these networks have enormous potential applicability, e.g., habitat monitoring, medical care, military surveillance or traffic control. Many protocols have been developed for Wireless Sensor Networks that try to overcome the constraints that characterize this type of networks. Ant-based routing protocols can add a significant contribution to assist in the maximisation of the network lifetime, but this is only possible by means of an adaptable and balanced algorithm that takes into account the Wireless Sensor Networks main restrictions. This paper presents a new Wireless Sensor Network routing protocol, which is based on the Ant Colony Optimization metaheuristic. The protocol was studied by simulation for several Wireless Sensor Network scenarios and the results clearly show that it minimises communication load and maximises energy savings.

In previous papers we proposed an algorithm for real parameter optimization called the Aggregation Pheromone System (APS). The APS replaces pheromone trails in traditional ACO with aggregation pheromones. The pheromone update rule is applied in a way similar to that of ACO. In this paper, we proposed an enhanced APS (eAPS), which uses a colony model with

units

. It allows a stronger exploitation of better solutions found and at the same time it can prevent premature stagnation of the search. Experimental results showed eAPS has higher performance than APS. It has also shown that the parameter settings for eAPS are more robust than for APS.

In this paper we present an estimation of distribution particle swarm optimization algorithm that borrows ideas from recent developments in ant colony optimization which can be considered an estimation of distribution algorithm. In the classical particle swarm optimization algorithm, particles exploit their individual memory to explore the search space. However, the swarm as a whole has no means to exploit its collective memory (represented by the array of previous best positions or pbests) to guide its search. This causes a re-exploration of already known bad regions of the search space, wasting costly function evaluations. In our approach, we use the swarm’s collective memory to probabilistically guide the particles’ movement towards the estimated promising regions in the search space. Our experiments show that this approach is able to find similar or better solutions than the canonical particle swarm optimizer with fewer function evaluations.

Data mining involves the automated process of finding patterns in data and has been a research topic for decades. Although very powerful data mining techniques exist to extract classification models from data, the techniques often infer counter-intuitive patterns or lack patterns that are logical for domain experts. The problem of consolidating the knowledge extracted from the data with the knowledge representing the experience of domain experts, is called the knowledge fusion problem. Providing a proper solution for this problem is a key success factor for any data mining application. In this paper, we explain how the AntMiner+ classification technique can be extended to incorporate such domain knowledge. By changing the environment and influencing the heuristic values, we can respectively limit and direct the search of the ants to those regions of the solution space that the expert believes to be logical and intuitive.

Assembly line balancing concerns the design of assembly lines for the manufacturing of products. In this paper we consider the time and space constrained simple assembly line balancing problem with the objective of minimizing the number of necessary work stations. This problem is denoted by TSALBP-1 in the literature. For tackling this problem we propose a Beam-ACO approach, which is an algorithm that results from hybridizing ant colony optimization with beam search. The experimental results show that our algorithm is a state-of-the-art metaheuristic for this problem.

This paper presents a novel boundary approach which is included as a constraint-handling technique in an ant colony algorithm. The necessity of approaching the boundary between the feasible and infeasible search space for many constrained optimization problems is a paramount challenge for every constraint-handling technique. Our proposed technique precisely focuses the search on the boundary region and can be either used alone or in combination with other constraint-handling techniques depending on the type and number of problem constraints. For validation purposes, an ant algorithm is adopted as our search engine. We compare our proposed approach with respect to constraint-handling techniques that are representative of the state-of-the-art in constrained evolutionary optimization using a set of standard test functions.

In this paper we analyse a previously introduced swarm intelligence control mechanism used for solving problems of robot path formation. We determine the impact of two probabilistic control parameters. In particular, the problem we consider consists in forming a path between two objects which an individual robot cannot perceive simultaneously.

Our experiments were conducted in simulation. We compare four different robot group sizes with up to 20 robots, and vary the difficulty of the task by considering five different distances between the objects which have to be connected by a path.

Our results show that the two investigated parameters have a strong impact on the behaviour of the overall system and that the optimal set of parameters is a function of group size and task difficulty. Additionally, we show that our system scales well with the number of robots.

We look at how the structure of social networks and the nature of social interactions affect the behaviour of Particle Swarms Optimisers. To this end, we propose a general model of communication and consensus which focuses on the effects of social interactions putting the details of the dynamics and the optimum seeking behaviour of PSOs into the background.

We present an algorithm for covering continuous domains by primitive robots whose only ability is to mark visited places with pheromone and to sense the level of the pheromone in their neighborhood. These pheromone marks can be sensed by all robots and thus provide a way for indirect communication between the robots. Apart from this, the robots have no means to communicate. Additionally they are memoryless, have no global information such as the domain map, own position, coverage percentage, etc. Despite the robots’ simplicity, we show that they are able to cover efficiently any connected domains, including non-planar ones.

Ant colony optimization algorithms are currently among the best performing algorithms for the quadratic assignment problem. These algorithms contain two main search procedures: solution construction by artificial ants and local search to improve the solutions constructed by the ants. Incremental local search is an approach that consists in re-optimizing partial solutions by a local search algorithm at regular intervals while constructing a complete solution. In this paper, we investigate the impact of adopting incremental local search in ant colony optimization to solve the quadratic assignment problem. Notwithstanding the promising results of incremental local search reported in the literature in a different context, the computational results of our new ACO algorithm are rather negative. We provide an empirical analysis that explains this failure.

In an ant society individuals coming from different groups (lines, strains) bear it own chemical identity and those individuals present discrimination capabilities between different chemical profiles. However, at the collective level these groups may cooperate and act together. To understand this apparent contradiction we have to keep in mind that amplification is the main component of many collective phenomena in social and gregarious insects. We use a model of food recruitment where each group of foragers have its own blend of pheromone trail that is partly recognized by the others groups. We found that a low level of recognition between signals is sufficient to produce a collaborative pattern between groups and that beyond a critical value of recognition. The aggregation of all the groups around the same food source is observed. Such collective response is a generic property of social phenomena governed by amplification processes.

Ant colony optimization (ACO) algorithms construct solutions each time starting from scratch, that is, from an

empty

solution. Differently from ACO algorithms, iterated greedy, another constructive stochastic local search method, starts the solution construction from

partial solutions

. In this paper, we examine the performance of a variation of

$\mathcal{MAX}$

-

$\mathcal{MIN}$

Ant System, one of the most successful ACO algorithms, that exploits this idea central to iterated greedy algorithms. We consider the quadratic assignment problem as a case-study, since this problem was also tackled in a closely related research to ours, the one on the usage of

external memory

in ACO. The usage of external memory resulted in ACO variants, where partial solutions are used to seed the solution construction. Contrary to previously reported results on external memory usage, our computational results are more pessimistic in the sense that starting the solution construction from partial solutions does not necessarily lead to improved performance when compared to state-of-the-art ACO algorithms.

In this paper, we study the cooperative transport of a heavy object by a group of robots towards a goal. We investigate the case in which robots have partial and noisy knowledge of the goal direction and can not perceive the goal itself. The robots have to coordinate their motion to apply enough force on the object to move it. Furthermore, the robots should share knowledge in order to collectively improve their estimate of the goal direction and transport the object as fast and as accurately as possible towards the goal.

We propose a bio-inspired mechanism of negotiation of direction that is fully distributed. Four different strategies are implemented and their performances are compared on a group of four real robots, varying the goal direction and the level of noise. We identify a strategy that enables efficient coordination of motion of the robots. Moreover, this strategy lets the robots improve their knowledge of the goal direction. Despite significant noise in the robots’ communication, we achieve effective cooperative transport towards the goal and observe that the negotiation of direction entails interesting properties of robustness.

The impact of the values of the most meaningful parameters on the behavior of

$\cal M\!AX\!$

–

$\cal MI\!N\!$

Ant System is analyzed. Namely, we take into account the number of ants, the evaporation rate of the pheromone, and the exponent values of the pheromone trail and of the heuristic measure in the random proportional rule. We propose an analytic approach to examining their impact on the speed of convergence of the algorithm. Some computational experiments are reported to show the practical relevance of the theoretical results.

It is often believed that the performance of ant system, and in general of ant colony optimization algorithms, depends somehow on the scale of the problem instance at hand. The issue has been recently raised explicitly [1] and the

hyper-cube framework

has been proposed to eliminate this supposed dependency.

In this paper, we show that although the internal state of ant system—that is, the

pheromone

matrix—depends on the scale of the problem instance under analysis, this does not affect the external behavior of the algorithm. In other words, for an appropriate initialization of the pheromone, the sequence of solutions obtained by ant system does not depend on the scale of the instance.

As a second contribution, the paper introduces a straightforward variant of ant system in which also the pheromone matrix is independent of the scale of the problem instance under analysis.

There are two reasons for parallelizing a metaheuristic if one is interested in performance: (i) given a fixed time to search, the aim is to increase the quality of the solutions found in that time; (ii) given a fixed solution quality, the aim is to reduce the time needed to find a solution not worse than that quality. In this article, we study the impact of communication when we parallelize a high-performing ant colony optimization (ACO) algorithm for the traveling salesman problem using message passing libraries. In particular, we examine synchronous and asynchronous communications on different interconnection topologies. We find that the simplest way of parallelizing the ACO algorithms, based on parallel independent runs, is surprisingly effective; we give some reasons as to why this is the case.

This paper presents a macrocell placement constraints and overlap removal methodology using particle swarm optimization (PSO). The authors adopted several techniques along with PSO as to avoid the floorplanning falling into the local minimum and to assist in finding out the global minimum. Our method can deal with various kinds of placement constraints, and consider them simultaneously. Experiments employing MCNC and GSRC benchmarks show the efficiency and robustness of our method for restricted placement and overlap removal obtained by the ability of exploring better solutions. The proposed approach exhibited rapid convergence and led to more optimal solutions than other related approaches, furthermore, it displayed efficient packing with all the constraints satisfied.

A central part of the rational drug development process is the prediction of the complex structure of a small ligand with a protein, the so-called protein-ligand docking problem, used in

virtual screening

of large databases and lead optimization. In the work presented here, we introduce a new docking algorithm called PLANTS (

P

rotein-

L

igand

ANTS

ystem), which is based on ant colony optimization. An artificial ant colony is employed to find a minimum energy conformation of the ligand in the protein’s binding site. We present the effectiveness of PLANTS for several parameter settings as well as a direct comparison to a state-of-the-art program called GOLD, which is based on a genetic algorithm. Last but not least, results for a virtual screening on the protein target factor Xa are presented.

This paper presents a novel glowworm metaphor based distributed algorithm that enables a minimalist mobile robot swarm to effectively split into subgroups, exhibit simultaneous taxis towards, and rendezvous at multiple source locations. The locations of interest could represent radiation sources such as nuclear and hazardous aerosol spills spread within an unknown environment. The glowworm algorithm is based on a glowworm swarm optimization (GSO) technique that finds multiple optima of multimodal functions. The algorithm is in the same spirit as the ant-colony optimization (ACO) and particle swarm optimization (PSO) algorithms, but with several significant differences. A key feature of the GSO algorithm is the use of an adaptive local-decision domain, which is used effectively to detect the multiple optimum locations of the multimodal function. We conduct sound source localization experiments, using a set of four wheeled robots (christened Glowworms), to validate the glowworm approach to the problem of multiple source localization. We also examine the behavior of the glowworm algorithm in the presence of uncertainty due to perceptional noise. A comparison with a gradient based approach reveals the superiority of the glowworm algorithm in coping with uncertainty.

This paper presents the application of the Ant Colony Optimization (ACO) meta-heuristic to a new NP-hard problem involving the replication of multi-quality database-driven web applications (DA s) by a large application service provider (ASP). The ASP must assign DA replicas to its network of heterogeneous servers so that user demand is satisfied at the desired quality level and replica update loads are minimized. Our ACO algorithm, AntDA , for solving the ASP’s replication problem has several novel or infrequently seen features: ants traverse a bipartite graph in both directions as they construct solutions, pheromone is used for traversing from one side of the bipartite graph to the other and back again, heuristic edge values change as ants construct solutions, and ants may sometimes produce infeasible solutions. Testing shows that the best results are achieved by using pheromone and heuristics to traverse the bipartite graph in both directions. Additionally, experiments show that AntDA outperforms several other solution methods.

CE-ants is a distributed, robust and adaptive swarm intelligence strategy for dealing with path management in communication networks. This paper focuses on various strategies for adjusting the overhead generated by the CE-ants as the state of the network changes. The overhead is in terms of number of management packets (ants) generated, and the adjustments are done by controlling the ant generation rate that controls the number ants traversing the network. The link state events considered are failure and restoration events. A simulation scenario compares restoration performance of rate adaptation in the source node with rate adaptation in the intermediate nodes close to the link state events. Implicit detection of failure events through monitoring ant parameters are considered. Results indicate that an implicit adjustment in the source node is a promising approach with respect to restoration time and the number of ants required.

In this paper we investigate the performance of pareto ant colony optimization (PACO) in solving a bi-objective permutation flowshop problem. We hybridize this technique by incorporating path relinking (PR) in four different ways. Several test instances are used to test the effectiveness of the different approaches. Computational results show that hybridizing PACO with PR improves the performance of PACO. The hybrid algorithms also show competitive results compared to other state of the art metaheuristics.

We have constructed a minimal cellular automaton model for simulating traffic on preexisting ant trails. Uni- as well as bidirectional trails are investigated and characteristic patterns in the distribution of workers on the trail are identified. Some of these patterns have already been observed empirically. They give rise to unusual flow characteristics which are also discussed. The question of possible functions of the observed traffic patterns and the resulting flow characteristics will be treated for simplified biological scenarios.

In this paper, a continuous Particle Swarm Optimization (

PSO

) algorithm is presented for the Uncapacitated Facility Location (

UFL

) problem. In order to improve the solution quality a local search is embedded to the

PSO

algorithm. It is applied to several benchmark suites collected from OR-library. The results are presented and compared to the results of two recent metaheuristic approaches, namely Genetic Algorithm(

GA

) and Evolutionary Simulated Annealing (

ESA

). It is concluded that the

PSO

algorithm is better than the compared methods and generates more robust results.

This paper proposes a direct application of Ant Colony Optimization to the function optimization problem in continuous domain. In the proposed algorithm, artificial ants construct solutions by selecting values for each variable randomly biased by a specific variable-related normal distribution, of which the mean and deviation values are represented by pheromone modified by ants according to the previous search experience. Some methods to avoid premature convergence, such as local search in different neighborhood structure, pheromone re-initialization and different solutions for pheromone intensification are incorporated into the proposed algorithm. Experimental setting of the parameters are presented, and the experimental results show the potential of the proposed algorithm in dealing with the function optimization problem of different characteristics.

This paper presents and implements an approach to parallel ACO algorithms. The principal idea is to make multiple ant colonies share and utilize only one pheromone matrix. We call it SHOP (SHaring One Pheromone matrix) approach. We apply this idea to the two currently best instances of ACO sequential algorithms (MMAS and ACS), and try to hybridize these two different ACO instances. We mainly describe how to design parallel ACS and MMAS based on SHOP. We present our computing results of applying our approach to solving 10 symmetric traveling salesman problems, and give comparisons with the relevant sequential versions under the fair computing environment. The experimental results indicate that SHOP-ACO algorithms perform overall better than the sequential ACO algorithms in both the computation time and solution quality.

Data clustering is one of important research topics of data mining. In this paper, we propose a new clustering algorithm based on ant colony optimization, called Ant Colony Optimization for Clustering (ACOC). At the core of the algorithm we use both the accumulated pheromone and the heuristic information, the distances between data objects and cluster centers of ants, to guide artificial ants to group data objects into proper clusters. This allows the algorithm to perform the clustering process more effectively and efficiently. Due to the nature of stochastic and population-based search, the ACOC can overcome the drawbacks of traditional clustering methods that easily converge to local optima. Experimental results show that the ACOC can find relatively good solutions.

The Capacitated Fixed Charge Location Problem (CFCLP) consists of selecting a set of facilities that must completely supply a set of customers at a minimum cost. The CFCLP is

NP-hard

thus solution methods are often obtained by using sophisticated techniques. However, if a set of facilities is known a priori then the CFCLP reduces to a transportation problem (TP). Although this can be used to derive solutions by randomly selecting sufficient facilities to be fixed open and noting any cost improvements, it is perceived as a poor technique that does not guarantee solutions near the optimal. This paper presents an adaptive sampling algorithm using Ant Colony Optimization (ACO). We hypothesize that random selection of facilities using ACO will generate at least near-optimal solutions for the CFCLP. Computational results for a series of standard benchmark problems are presented which appear to confirm this hypothesis.

This paper studies the open vehicle routing problem (OVRP), in which the vehicles do not return to the starting depot after serving the last customers or, if they do, they must make the same trip in the reverse order. We present an ant colony system hybridized with local search for solving the OVRP (ACS-OVRP). Additionally, a Post-Optimization procedure is incorporated in the proposed algorithm to further improve the best-found solutions. The computational results of ACS-OVRP compared to those of other algorithms are reported, which indicate that the ACS-OVRP is another efficient algorithm for solving the OVRP.

In this article a method for color reduction based on ant colony algorithm is presented. Generally color reduction involves two steps: choosing a proper palette and mapping colors to this palette. This article is about the first step. Using ant colony algorithm, pixel clusters are formed based on their colors and neighborhood information to make final palette. A comparison between the results of the proposed method and some other methods is presented. There are some parameters in the proposed method which can be set based on user needs and priorities. This increases the flexibility of the method.

Ant colony optimization has been one of the most promising meta-heuristics since its appearance in early 1990s but it is specialized in discrete space optimization problems. To explore the utility of ACO in the filed of continuous problems, this paper proposes an orthogonal search embedded ACO (OSEACO) algorithm. By generating some grids in the search space and embedding an orthogonal search scheme into ACO, the search space is learned much more comprehensively with only few computation efforts consumed. Hence, solutions are obtained in higher precision. Some adaptive strategies are also developed to prevent the algorithm from trapping in local optima as well as to improve its performance. Moreover, the effectiveness of this algorithm is demonstrated by experimental results on 9 diverse test functions for it is able to obtain near-optimal solutions in all cases.

Many emergency situations involve injured people who need medical help and evacuation to a safe area. Usually there is not enough time to provide medical help to all the victims. So medical doctors have to make choices to help as much victims as possible, taking care of the distribution of victims in the crisis area and the priority of the victims related to the severity of the injuries. This paper describes an ant colony optimization algorithm to route medical doctors along the victims in a crisis area. We tested the algorithm in a simulated crisis environment. Two different routing strategies were implemented and compared with our algorithm.

They appeared in our life some years ago with the awakening of the PC and now the are everywhere : computers have become ubiquitous and, almost, irreplaceable. Classical ways of creating, managing and exchanging information have been progressively replaced by electronic means. In spite of this plebiscite, computer supported collaborative work (CSCW) softwares can be blamed for requiring the user to do an effort to use them. This paper describes an artificial ant based autonomous information dissemination algorithm. It constitutes the communication layer of our framework PIAF (“Personal Intelligent Agent Framework”) intended to help users transparently sharing information. The algorithm uses message gossiping strategy to transfer information items between users.

Stigmergy is a form of indirect interaction for coordination and communication purposes that can be found in many swarm systems. In this paper we present a tiled coprocessor for computation-intensive applications that explicitly exploits stigmergy to achieve adaptability avoiding the usual time-consuming handshakes involved in direct interactions. This adaptability, without any centralized control, directly implies architectural scalability at design time, flexibility in multitasking environment, adaptive load balancing and fault-tolerance at run-time. A CMOS 0.13

μ

m implementation of such architecture for simple array processing operations is presented and evaluated. Obtained results show the potentiality of the proposed approach.

This paper describes a distributed algorithm for solving the shortest path problem with a swarm of JASMINE micro-robots. Each robot is only connected via infra-red communication with its neighbours. Based on local information exchange and some simple rules the swarm manages to find the shortest path (shortest path in the number of robots on the path) in a labyrinth with dead-ends and cycles. The full algorithm and simulation results are presented in this paper.

The paper presents highlights of a doctoral research addressed to model and solve the problem of preventive maintenance scheduling of fleets of vehicles. The problem is formulated and several Ant Colony based approaches were proposed and tested considering different instances of the maintenance scheduling problem, including applications related to the preventive maintenance of an aircraft fleet belonging to the Brazilian Air Force. The most successful approach has shown to be the one based on ACS with a specific local search procedure and inspired on the application of ACO to the timetabling problem. Details on the problem formulation, on the proposed approaches, and on the performed tests and applications are presented.

Inverse problems in ocean acoustics are generally solved by means of matched field processing in combination with metaheuristic global search algorithms. Solutions that describe acoustical properties of the bed and subbottom in a shallow water environment are typically approximations that require uncertainty analysis. This work compares Ant Colony Optimization with other metaheuristics for geoacoustic inversion, particularly Genetic Algorithms. It is demonstrated that a

$\mathcal{MAX}$

-

$\mathcal{MIN}$

Ant System can find good estimates and provide uncertainty analysis. In addition, the algorithm can easily be tuned, but proper tuning does not guarantee that every run will converge given a limited processing time. Another concern is that a single optimization run may find a solution while there is no clear indication on the accuracy. Both issues can be solved when probability distributions are based on parallel

$\mathcal{MAX-MIN}$

Ant System runs.

Ant colony optimisation is a constructive metaheuristic that successively builds solutions from problem-specific components. A parameterised model known as pheromone—an analogue of the trail pheromones used by real ants—is used to learn which components should be combined to produce good solutions. In the majority of the algorithm’s applications a single parameter from the model is used to influence the selection of a single component to add to a solution. Such a model can be described as first order. Higher order models describe relationships between several components in a solution, and may arise either by contriving a model that describes subsets of components from a first order model or because the characteristics of solutions modelled naturally relate subsets of components. This paper introduces a simple framework to describe the application of higher order models as a tool to understanding common features of existing applications. The framework also serves as an introduction to those new to the use of such models. The utility of higher order models is discussed with reference to empirical results in the literature.

In this article, we study hybrid Particle Swarm Optimization (PSO) algorithms for continuous optimization. The algorithms combine a PSO algorithm with either the Nelder-Mead-Simplex or Powell’s Direction-Set local search methods. Local search is applied each time the PSO part meets some convergence criterion. Our experimental results for test functions with up to 100 dimensions indicate that the usage of the iterative improvement algorithms can strongly improve PSO performance but also that the preferable choice of which local search algorithm to apply depends on the test function. The results also suggest that another main contribution of the local search is to make PSO algorithms more robust with respect to their parameter settings.

This paper proposes a Binary Ant Colony Optimization applied to constrained optimization problems with binary solution structure. Due to its simple structure, the convergence status of the proposed algorithm can be monitored through the distribution of pheromone in the solution space, and the probability of solution improvement can be in some way controlled by the maintenance of pheromone. The successful implementations to the binary function optimization problem and the multidimensional knapsack problem indicate the robustness and practicability of the proposed algorithm.

For solving combinatorial optimisation problems, exact methods accurately exploit the structure of the problem but are tractable only up to a certain size; approximation or heuristic methods are tractable for very large problems but may possibly be led into a bad solution. A question that arises is, From where can we obtain knowledge of the problem structure via exact methods that can be exploited on large-scale problems by heuristic methods? We present a framework that allows the exploitation of existing techniques and resources to integrate such structural knowledge into the Ant Colony System metaheuristic, where the structure is determined through the notion of kernelization from the field of parameterized complexity. We give experimental results using vertex cover as the problem instance, and show that knowledge of this type of structure improves performance beyond previously defined ACS algorithms.

In this paper, a discrete particle swarm optimization (DPSO) algorithm is presented to solve the single machine total earliness and tardiness penalties with a common due date. A modified version of HRM heuristic presented by Hino et al. in [1], here we call it MHRM, is also presented to solve the problem. In addition, the DPSO algorithm is hybridized with the iterated local search (ILS) algorithm to further improve the solution quality. The performance of the proposed DPSO algorithm is tested on 280 benchmark instances ranging from 10 to 1000 jobs from the OR Library. The computational experiments showed that the proposed DPSO algorithm has generated better results, in terms of both percentage relative deviations from the upper bounds in Biskup and Feldmann and computational time, than Hino et al. [1].

Support vector machines, especially when using radial basis kernels, have given good results in the classification of different volatile compounds. We can achieve a feature extraction method adjusting the parameters of a modified radial basis kernel, giving more importance to those features that are important for classification proposes. However, the function that has to be minimized to find the best scaling factors is not derivable and has multiple local minima. In this work we propose to adapt the ideas of the ant colony optimization method to find an optimal value of the kernel parameters.

From October 2005 to August 2006, the museum of natural history of the city of Tours is displaying a temporary exhibition designed by the members of the “handicap and new technologies” research team from the computer science lab of the University of Tours. This paper describes this exhibition, the goals and means that have been involved to popularize recent researches in the field of artificial insects. Different robots and computer simulations have been used to explain how our community builds new paradigms inspired by insect models and swarm intelligence concepts.

Production scheduling problems such as the job shop consist of a collection of operations (grouped into jobs) that must be scheduled for processing on different machines. Typical ant colony optimisation applications for these problems generate solutions by constructing a permutation of the operations, from which a deterministic algorithm can generate the actual schedule. This paper considers an alternative approach in which each machine is assigned a dispatching rule, which heuristically determines the order of operations on that machine. This representation creates a substantially smaller search space that likely contains good solutions. The performance of both approaches is compared on a real-world job shop scheduling problem in which processing times and job due dates are modelled with fuzzy sets. Results indicate that the new approach produces better solutions more quickly than the traditional approach.

The exam timetabling problem faces the problem of scheduling exams within a limited number of available periods. The main objective is to balance out student’s workload by distributing the exams evenly within the planning horizon. Ant colony approaches have been proven to be a powerful solution approach for various combinatorial optimization problems. In this paper a Max-Min and a ANTCOL approach will be presented. Its performance is compared with other approaches presented in the literature and with modified graph coloring algorithms.

In 1990, Deneubourg et al. [1] developed the first ant clustering algorithm based on mimicking corpse piling process of ants. In his algorithm, an ant picks up and drops the data items based on the similarity of ant’s local neighborhoods. Labroche et al. [2] developed a different ant algorithm, AntClust, based on chemical odor and some behavioral rules of ants.

This reasearch studies ACO algorithms for the switch engine scheduling in a Railroad Yard. The cars are moved individually or grouped into blocks by a set of locomotives called switch engines which are dedicated to moving around the cars in the yard. The need for moving comes from the fact that the arriving trains are disassembled into blocks of cars, undergo some operations like, loading, unloading and cleaning and finally are assembled into a new train. Each moving request is called a switch order. The decision of which switch engine will execute which switch order and the sequence of that executution is the core of our problem. The optimization of this schedule reduces the overall operational costs of the yard as weel as the time to assemble new trains, thus leading to a more productive railroad system.

Although ACO has been proved to be an efficient and versatile tool for combinational optimization problems [1,2], it cannot deal with continuous optimization problems directly. Therefore, there are only a few studies on ACO [3] for continuous optimization. This paper presents a novel ACO algorithm (CACO-DE) for continuous optimization based on discrete encoding, which is quite different from other ant methods.

Throughout the history of research, some of the most innovative and useful discoveries have arisen from a fusion of two or more seemingly unrelated fields of study; a characteristic of some method or process is enfused into a completely disjoint technique, and the resulting creation exhibits superior behavior. Some common examples include simulated annealing modeled after the annealing process in physics, Ant Colony Optimization modeled after the behavior of social insects, and the Particle Swarm Optimization algorithm modeled after the patterns of flocking birds.

In the Single Source Capacitated Facility Location Problem (SSCFLP),

n

customers and

m

potential facility locations are given. Each customer

i

has an associated demand

d

i

that must be supplied only by a single facility

j

. The cost

c

ij

is the unit shipping cost between customer

i

and facility

j

. The capacity of each facility

s

j

is known, as well as the fixed charge

f

j

incurred whenever facility

j

is opened. The objective of the problem is to locate a number of facilities that serve a set of customers at minimum cost.

We formulate a non linear optimization problem to find the optimal sink node position for a given wireless sensor networks (WSN) region where sensor nodes generate different amounts of data to send to the sink node [1]. The problem is NP-hard in general and we use the particle swarm optimization technique to solve the optimization problem [2].

Swarm Intelligent (SI) algorithms draw their inspiration from the interaction of individuals of social organisms. One such algorithm, Ant Colony Optimization (ACO) [1], utilizes the foraging behavior of ants to solve combinatorial optimization problems. Although ACO performs well in a static environment, it has been pointed out that ACO does not perform as well as other heuristics in dynamic situations such as routing. This paper proposes a new algorithm, entitled Evolutionary Ant Colony Optimization (EACO), that combines ACO with elements of Genetic Algorithms (GA). By adding evolution, the EACO algorithm allows the individual ants to develop their own characteristics, thereby removing the homogeneity inherent within ACO. Our results demonstrate the potential of this approach in a dynamic environment.

The particle swarm algorithm [1] contains elements which map fairly strongly to the

group-foraging

problem in behavioural ecology: its continuous equations of motion include concepts of social attraction and communication between individuals, two of the general requirements for grouping behaviour [2]. Despite its socio-biological background, the particle swarm algorithm has rarely been applied to biological problems, largely remaining a technique used in classical optimisation problems. In this paper [3], we show how some simple adaptions to the standard algorithm can make it well suited for the foraging problem.

The facility layout problems in today’s batch-to-mass manufacturing systems have gained a whole new face and popularity due to the requirements of mass customization. Facilities layout problem is still the most popular application for the Quadratic Assignment Problem [1]. The facility layout model addressed in this study is also known as quadratic assignment problem and formulated as integer linear programming formulation [2]. PSO performs a population-based search which emulates the social behavior of bird flocking and fish schooling [3]. In PSO, candidate solutions, called particles, search through the problem space by following the current optimal particles. Locations of particles in the search space are determined by their positions and velocities.

Multi-Agent Systems (MAS) are extensively used as a tool for simulation of dynamic systems. Geosimulation is an urban phenomena approach that uses the multi-agent methodology to simulate discrete, dynamic, and event-oriented systems. Our focus in this paper is to use self-organization, specially strategies

inspired

by solutions from Swarm Intelligence, as well as the idea of social networks, and demonstrate their effect on learning in geosimulation agents.

Several traffic control approaches address the problem of reducing traffic jams. A class of them deals with coordination of traffic lights to allow vehicles traveling in a given direction to pass an arterial without stopping. However, in cities where the business centers are no longer located exclusively downtown, this approach is not appropriate: simple offline optimization of the synchronization in

one arterial

alone cannot cope with changing traffic patterns.

We propose an extended ACO algorithm for the Maximum Edge Disjoint Paths (MEDP) problem. In this problem we want to satisfy the largest possible number of request for disjoint paths on a given graph topology. We first proposed our approach in [1]. In that paper a proof of concept was given on a number of quite small graphs. Now we build on that approach and use existing ACO features to develop an algorithm capable of obtaining good results on a set of MEDP benchmark problems.