In this chapter we describe evolutionary algorithms (EAs) for constraint handling. Constraint han... more In this chapter we describe evolutionary algorithms (EAs) for constraint handling. Constraint handling is not straightforward in an EA because the search operators mutation and recombination are "blind" to constraints. Hence, there is no guarantee that if the parents satisfy some constraints the offspring will satisfy them as well. This suggests that the presence of constraints in a problem makes EAs intrinsically unsuited to solve this problem. This should especially hold when the problem does not contain an objective function to be optimized, but only constraints -the category of constraint satisfaction problems. A survey of related literature, however, indicates that there are quite a few successful attempts to evolutionary constraint satisfaction. Based on this survey, we identify a number of common features in these approaches and arrive at the conclusion that EAs can be effective constraint solvers when knowledge about the constraints is incorporated either into the genetic operators, in the fitness function, or in repair mechanisms. We conclude by considering a number of key questions on research methodology.
Proceedings of the 2010 Winter Simulation Conference, 2010
Multi-Agent Systems (MAS) are increasingly used to solve larger and more complex problems. To pro... more Multi-Agent Systems (MAS) are increasingly used to solve larger and more complex problems. To provide the computational resources needed to do this, MAS are increasingly distributed over multiple computational platforms. Different approaches for distributing MAS have been proposed over the years. One problem remains central whichever approach is used: how to translate MAS behaviour into a format suitable for the distribution approach used. In this paper we describe the Agent Distributed Shared Memory Interface (ADSMI), an interface between a general MAS description and the PDES-MAS platform as an implementation of a Distributed Shared Memory (DSM) system for distributed MAS simulations. The ADSMI provides a translation of MAS behaviour into event-interactions with PDES-MAS, as well as functionality for handling time progressing in the MAS, and a messaging mechanism between agents.
2011 Fifth International Conference on Innovative Mobile and Internet Services in Ubiquitous Computing, 2011
As much as ubiquitous computing systems are already claimed to exist in the real world, further d... more As much as ubiquitous computing systems are already claimed to exist in the real world, further development of these systems still pose challenges to computer science that are still quite beyond the state of the art. Two challenges stand out in particular: the complexity of next-generation ubiquitous computing systems, and their inherent scalability issues. This paper aims to establish that agent-based modelling provides a powerful tool in tackling these issues. As an example of a practical solution, readily available, this paper highlights the distributed agent-based simulation infrastructure PDES-MAS as particularly suited for the task. Using the PDES-MAS infrastructure, designers, developers, and builders of next-generation ubiquitous computing systems can, through an iterative agent-based simulation process, gain the required knowledge and information about these systems, without having precede to deployment of the system itself.
2005 IEEE Congress on Evolutionary Computation, 2005
We study a selected group of hybrid EAs for solving CSPs, consisting of the best performing EAs f... more We study a selected group of hybrid EAs for solving CSPs, consisting of the best performing EAs from the literature. We investigate the contribution of the evolutionary component to their performance by comparing the hybrid EAs with their "de-evolutionarised" variants. The experiments show that "de-evolutionarising" can increase performance, in some cases doubling it. Considering that the problem domain and the algorithms are arbitrarily selected from the "memetic niche", it seems likely that the same effect occurs for other problems and algorithms. Therefore, our conclusion is that after designing and building a memetic algorithm, one should perform a verification by comparing this algorithm with its "de-evolutionarised" variant.
2011 IEEE/ACM 15th International Symposium on Distributed Simulation and Real Time Applications, 2011
Groundwater is like dark matter -we know very little apart from the fact that it is hugely import... more Groundwater is like dark matter -we know very little apart from the fact that it is hugely important. Given the scarcity of data, mathematical modelling can come to the rescue but existing groundwater models are mainly restricted to simulate the transport and degradation of contaminants on the scale of whole contaminated field sites by averaging out the effect of spatial heterogeneity on the availability of the pollutant to the degrading organisms. These coarse-scale mean-field models therefore tend to rely on fitting to data rather than being predictive. Also, they are less suited to incorporate spatial variability and non-linear kinetics and feedbacks. We propose to solve the two mutually exacerbating problems of environmental patchiness and data scarcity by developing a flexible and robust distributed simulation framework that uses an ensemble of small scale simulations running on different processors/computers to scale-up, i.e. to feed the effect of small-scale patchiness into a concurrent site-scale simulation of the dynamics of groundwater pollutant degradation. Our scaling approach solves problem #1 by simulating dynamics also on the small scale where some of the patchiness resides, and problem #2 by enabling rigorous validation of our small-scale model and scaling approach with laboratory data, which are high quality at low cost.
IEEE Transactions on Evolutionary Computation, 2003
Constraint handling is not straightforward in evolutionary algorithms (ea) since the usual search... more Constraint handling is not straightforward in evolutionary algorithms (ea) since the usual search operators, mutation and recombination, are 'blind' to constraints. Nevertheless, the issue is highly relevant, for many challenging problems involve constraints. Over the last decade numerous eas for solving constraint satisfaction problems (csp) have been introduced and studied on various problems. The diversity of approaches and the variety of problems used to study the resulting algorithms prevents a fair and accurate comparison of these algorithms. This paper aligns related work by presenting a concise overview and an extensive performance comparison of all these eas on a systematically generated test suite of random binary csps. The random problem instance generator is based on a theoretical model that fixes deficiencies of models and respective generators that have been formerly used in the Evolutionary Computing (ec) field.
Game developers have always striven to create better and more complex games. The use of artificia... more Game developers have always striven to create better and more complex games. The use of artificial intelligence (AI) in games is increasing in game development to this end. However there appears to be a mismatch of implications and expectations between game developers and AI researchers of what AI can bring to games, mostly due to differences of interpretation of the problems involved. To overcome this we present a taxonomy to aid knowledge transfer between the two groups, including elements from environments, game theory, and information theory. We also identify other concepts of games and AI that could be important to both sides.
ACM Transactions on Modeling and Computer Simulation, 2013
Range-Query is an important associative form of data access in distributed simulations and Distri... more Range-Query is an important associative form of data access in distributed simulations and Distributed Virtual Environments. This paper discusses the problem of Range-Queries in the context of distributed simulation of multiagent systems. An algorithm is presented for performing instantaneous Queries within an optimistic synchronisation framework and in the presence of dynamic migration of the simulation state. A quantitative evaluation of the effectiveness of the algorithm under different conditions is also presented.
Although historical studies are frequently perceived as clear narratives defined by a series of f... more Although historical studies are frequently perceived as clear narratives defined by a series of fixed events; in reality, even where critical historical events may be identified, historic documentation frequently lacks corroborative detail to support verifiable interpretation. Consequently, interpretation rarely rises above the level of unproven assertion and is rarely tested against a range of evidence. Agent-based simulation can provide an opportunity to break these cycles of academic claim and counter-claim.
Traditionally, history has frequently emphasized the role of the ‘Great Man or Woman’, who may ac... more Traditionally, history has frequently emphasized the role of the ‘Great Man or Woman’, who may achieve greatness, or notoriety, through the consequences of their decisions. More problematic is the historical treatment of the mass of the population. Agent-based modelling is a computer simulation technique that can not only help identify key interactions that contribute to large scale patterns but also add detail to our understanding of the effects of all contributors to a system, not just those at the top. The Medieval Warfare on the Grid project has been using agent-based models to examine the march of the Byzantine army across Anatolia to Manzikert in AD 1071. This article describes the movement model used to simulate the army and the historical sources on which it was based. It also explains why novel route planning algorithms were required in order to surmount problems with standard solutions.
This paper introduces a hybrid Tabu Search -Evolutionary Algorithm for solving the binary constra... more This paper introduces a hybrid Tabu Search -Evolutionary Algorithm for solving the binary constraint satisfaction problem, called CTLEA. A continuation of an earlier introduced algorithm, called the STLEA, the CTLEA replaces the earlier compound label tabu list with a conflict tabu list. Extensive experimental fine-tuning of parameters was performed to optimise the performance of the algorithm on a commonly used test-set. Compared to the performance of the earlier STLEA, and benchmark algorithms, the CTLEA outperforms the former, and approaches the later.
Recently, peer-to-peer networks have been proposed as the underlying architecture of large scale ... more Recently, peer-to-peer networks have been proposed as the underlying architecture of large scale distributed social agent simulations. A number of problems arise when grid landscapes are used to represent the landscape in these simulations, primarily because, in a peer-to-peer network, the landscape has to be handled collectively by the nodes of the network. Two basic agent actions are identified as central to these problems: look and move. A solution to these problems is provided in which the network maintains a move-buffer and a look-index. Both solutions are configurable by the user of the simulation and provide a trade-off between the scalability of the system, the consistency of the information stored in the system, and the efficiency of the system.
The paper introduces a hybrid Tabu Search-Evolutionary Algorithm for solving the constraint satis... more The paper introduces a hybrid Tabu Search-Evolutionary Algorithm for solving the constraint satisfaction problem, called STLEA. Extensive experimental fine-tuning of parameters of the algorithm was performed to optimise the performance of the algorithm on a commonly used test-set. The performance of the STLEA was then compared to the best known evolutionary algorithm and benchmark deterministic and non-deterministic algorithms. The comparison shows that the STLEA improves on the performance of the best known evolutionary algorithm but can not achieve the efficiency of the deterministic algorithms.
This paper introduces a hybrid Tabu Search - Evolutionary Algorithm for solving the binary constr... more This paper introduces a hybrid Tabu Search - Evolutionary Algorithm for solving the binary constraint satisfaction problem, called CTLEA. A continuation of an earlier introduced algorithm, called the STLEA, the CTLEA replaces the earlier compound label tabu list with a conflict tabu list. Extensive experimental fine-tuning of parameters was performed to optimise the performance of the algorithm on a commonly
The paper introduces a hybrid Tabu Search-Evolutionary Algorithm for solving the constraint satis... more The paper introduces a hybrid Tabu Search-Evolutionary Algorithm for solving the constraint satisfaction problem, called STLEA. Extensive experimental fine-tuning of parameters of the algorithm was performed to optimise the performance of the algorithm on a commonly used test-set. The performance of the STLEA was then compared to the best known evolutionary algorithm and benchmark deterministic and non-deterministic algorithms. The comparison
Journal of Artificial Societies and Social Simulation, Mar 31, 2006
The NewTies project is implementing a simulation in which societies of agents are expected to de-... more The NewTies project is implementing a simulation in which societies of agents are expected to de-velop autonomously as a result of individual, population and social learning. These societies are expected to be able to solve environmental challenges by acting collectively. The challenges are in-tended to be analogous to those faced by early, simple, small-scale human societies. This report on work in progress outlines the major features of the system as it is currently conceived within the project, including the design of the agents, the environment, the mechanism for the evolution of language and the peer-to-peer infrastructure on which the simulation runs.
In this chapter we describe evolutionary algorithms (EAs) for constraint handling. Constraint han... more In this chapter we describe evolutionary algorithms (EAs) for constraint handling. Constraint handling is not straightforward in an EA because the search operators mutation and recombination are "blind" to constraints. Hence, there is no guarantee that if the parents satisfy some constraints the offspring will satisfy them as well. This suggests that the presence of constraints in a problem makes EAs intrinsically unsuited to solve this problem. This should especially hold when the problem does not contain an objective function to be optimized, but only constraints -the category of constraint satisfaction problems. A survey of related literature, however, indicates that there are quite a few successful attempts to evolutionary constraint satisfaction. Based on this survey, we identify a number of common features in these approaches and arrive at the conclusion that EAs can be effective constraint solvers when knowledge about the constraints is incorporated either into the genetic operators, in the fitness function, or in repair mechanisms. We conclude by considering a number of key questions on research methodology.
Proceedings of the 2010 Winter Simulation Conference, 2010
Multi-Agent Systems (MAS) are increasingly used to solve larger and more complex problems. To pro... more Multi-Agent Systems (MAS) are increasingly used to solve larger and more complex problems. To provide the computational resources needed to do this, MAS are increasingly distributed over multiple computational platforms. Different approaches for distributing MAS have been proposed over the years. One problem remains central whichever approach is used: how to translate MAS behaviour into a format suitable for the distribution approach used. In this paper we describe the Agent Distributed Shared Memory Interface (ADSMI), an interface between a general MAS description and the PDES-MAS platform as an implementation of a Distributed Shared Memory (DSM) system for distributed MAS simulations. The ADSMI provides a translation of MAS behaviour into event-interactions with PDES-MAS, as well as functionality for handling time progressing in the MAS, and a messaging mechanism between agents.
2011 Fifth International Conference on Innovative Mobile and Internet Services in Ubiquitous Computing, 2011
As much as ubiquitous computing systems are already claimed to exist in the real world, further d... more As much as ubiquitous computing systems are already claimed to exist in the real world, further development of these systems still pose challenges to computer science that are still quite beyond the state of the art. Two challenges stand out in particular: the complexity of next-generation ubiquitous computing systems, and their inherent scalability issues. This paper aims to establish that agent-based modelling provides a powerful tool in tackling these issues. As an example of a practical solution, readily available, this paper highlights the distributed agent-based simulation infrastructure PDES-MAS as particularly suited for the task. Using the PDES-MAS infrastructure, designers, developers, and builders of next-generation ubiquitous computing systems can, through an iterative agent-based simulation process, gain the required knowledge and information about these systems, without having precede to deployment of the system itself.
2005 IEEE Congress on Evolutionary Computation, 2005
We study a selected group of hybrid EAs for solving CSPs, consisting of the best performing EAs f... more We study a selected group of hybrid EAs for solving CSPs, consisting of the best performing EAs from the literature. We investigate the contribution of the evolutionary component to their performance by comparing the hybrid EAs with their "de-evolutionarised" variants. The experiments show that "de-evolutionarising" can increase performance, in some cases doubling it. Considering that the problem domain and the algorithms are arbitrarily selected from the "memetic niche", it seems likely that the same effect occurs for other problems and algorithms. Therefore, our conclusion is that after designing and building a memetic algorithm, one should perform a verification by comparing this algorithm with its "de-evolutionarised" variant.
2011 IEEE/ACM 15th International Symposium on Distributed Simulation and Real Time Applications, 2011
Groundwater is like dark matter -we know very little apart from the fact that it is hugely import... more Groundwater is like dark matter -we know very little apart from the fact that it is hugely important. Given the scarcity of data, mathematical modelling can come to the rescue but existing groundwater models are mainly restricted to simulate the transport and degradation of contaminants on the scale of whole contaminated field sites by averaging out the effect of spatial heterogeneity on the availability of the pollutant to the degrading organisms. These coarse-scale mean-field models therefore tend to rely on fitting to data rather than being predictive. Also, they are less suited to incorporate spatial variability and non-linear kinetics and feedbacks. We propose to solve the two mutually exacerbating problems of environmental patchiness and data scarcity by developing a flexible and robust distributed simulation framework that uses an ensemble of small scale simulations running on different processors/computers to scale-up, i.e. to feed the effect of small-scale patchiness into a concurrent site-scale simulation of the dynamics of groundwater pollutant degradation. Our scaling approach solves problem #1 by simulating dynamics also on the small scale where some of the patchiness resides, and problem #2 by enabling rigorous validation of our small-scale model and scaling approach with laboratory data, which are high quality at low cost.
IEEE Transactions on Evolutionary Computation, 2003
Constraint handling is not straightforward in evolutionary algorithms (ea) since the usual search... more Constraint handling is not straightforward in evolutionary algorithms (ea) since the usual search operators, mutation and recombination, are 'blind' to constraints. Nevertheless, the issue is highly relevant, for many challenging problems involve constraints. Over the last decade numerous eas for solving constraint satisfaction problems (csp) have been introduced and studied on various problems. The diversity of approaches and the variety of problems used to study the resulting algorithms prevents a fair and accurate comparison of these algorithms. This paper aligns related work by presenting a concise overview and an extensive performance comparison of all these eas on a systematically generated test suite of random binary csps. The random problem instance generator is based on a theoretical model that fixes deficiencies of models and respective generators that have been formerly used in the Evolutionary Computing (ec) field.
Game developers have always striven to create better and more complex games. The use of artificia... more Game developers have always striven to create better and more complex games. The use of artificial intelligence (AI) in games is increasing in game development to this end. However there appears to be a mismatch of implications and expectations between game developers and AI researchers of what AI can bring to games, mostly due to differences of interpretation of the problems involved. To overcome this we present a taxonomy to aid knowledge transfer between the two groups, including elements from environments, game theory, and information theory. We also identify other concepts of games and AI that could be important to both sides.
ACM Transactions on Modeling and Computer Simulation, 2013
Range-Query is an important associative form of data access in distributed simulations and Distri... more Range-Query is an important associative form of data access in distributed simulations and Distributed Virtual Environments. This paper discusses the problem of Range-Queries in the context of distributed simulation of multiagent systems. An algorithm is presented for performing instantaneous Queries within an optimistic synchronisation framework and in the presence of dynamic migration of the simulation state. A quantitative evaluation of the effectiveness of the algorithm under different conditions is also presented.
Although historical studies are frequently perceived as clear narratives defined by a series of f... more Although historical studies are frequently perceived as clear narratives defined by a series of fixed events; in reality, even where critical historical events may be identified, historic documentation frequently lacks corroborative detail to support verifiable interpretation. Consequently, interpretation rarely rises above the level of unproven assertion and is rarely tested against a range of evidence. Agent-based simulation can provide an opportunity to break these cycles of academic claim and counter-claim.
Traditionally, history has frequently emphasized the role of the ‘Great Man or Woman’, who may ac... more Traditionally, history has frequently emphasized the role of the ‘Great Man or Woman’, who may achieve greatness, or notoriety, through the consequences of their decisions. More problematic is the historical treatment of the mass of the population. Agent-based modelling is a computer simulation technique that can not only help identify key interactions that contribute to large scale patterns but also add detail to our understanding of the effects of all contributors to a system, not just those at the top. The Medieval Warfare on the Grid project has been using agent-based models to examine the march of the Byzantine army across Anatolia to Manzikert in AD 1071. This article describes the movement model used to simulate the army and the historical sources on which it was based. It also explains why novel route planning algorithms were required in order to surmount problems with standard solutions.
This paper introduces a hybrid Tabu Search -Evolutionary Algorithm for solving the binary constra... more This paper introduces a hybrid Tabu Search -Evolutionary Algorithm for solving the binary constraint satisfaction problem, called CTLEA. A continuation of an earlier introduced algorithm, called the STLEA, the CTLEA replaces the earlier compound label tabu list with a conflict tabu list. Extensive experimental fine-tuning of parameters was performed to optimise the performance of the algorithm on a commonly used test-set. Compared to the performance of the earlier STLEA, and benchmark algorithms, the CTLEA outperforms the former, and approaches the later.
Recently, peer-to-peer networks have been proposed as the underlying architecture of large scale ... more Recently, peer-to-peer networks have been proposed as the underlying architecture of large scale distributed social agent simulations. A number of problems arise when grid landscapes are used to represent the landscape in these simulations, primarily because, in a peer-to-peer network, the landscape has to be handled collectively by the nodes of the network. Two basic agent actions are identified as central to these problems: look and move. A solution to these problems is provided in which the network maintains a move-buffer and a look-index. Both solutions are configurable by the user of the simulation and provide a trade-off between the scalability of the system, the consistency of the information stored in the system, and the efficiency of the system.
The paper introduces a hybrid Tabu Search-Evolutionary Algorithm for solving the constraint satis... more The paper introduces a hybrid Tabu Search-Evolutionary Algorithm for solving the constraint satisfaction problem, called STLEA. Extensive experimental fine-tuning of parameters of the algorithm was performed to optimise the performance of the algorithm on a commonly used test-set. The performance of the STLEA was then compared to the best known evolutionary algorithm and benchmark deterministic and non-deterministic algorithms. The comparison shows that the STLEA improves on the performance of the best known evolutionary algorithm but can not achieve the efficiency of the deterministic algorithms.
This paper introduces a hybrid Tabu Search - Evolutionary Algorithm for solving the binary constr... more This paper introduces a hybrid Tabu Search - Evolutionary Algorithm for solving the binary constraint satisfaction problem, called CTLEA. A continuation of an earlier introduced algorithm, called the STLEA, the CTLEA replaces the earlier compound label tabu list with a conflict tabu list. Extensive experimental fine-tuning of parameters was performed to optimise the performance of the algorithm on a commonly
The paper introduces a hybrid Tabu Search-Evolutionary Algorithm for solving the constraint satis... more The paper introduces a hybrid Tabu Search-Evolutionary Algorithm for solving the constraint satisfaction problem, called STLEA. Extensive experimental fine-tuning of parameters of the algorithm was performed to optimise the performance of the algorithm on a commonly used test-set. The performance of the STLEA was then compared to the best known evolutionary algorithm and benchmark deterministic and non-deterministic algorithms. The comparison
Journal of Artificial Societies and Social Simulation, Mar 31, 2006
The NewTies project is implementing a simulation in which societies of agents are expected to de-... more The NewTies project is implementing a simulation in which societies of agents are expected to de-velop autonomously as a result of individual, population and social learning. These societies are expected to be able to solve environmental challenges by acting collectively. The challenges are in-tended to be analogous to those faced by early, simple, small-scale human societies. This report on work in progress outlines the major features of the system as it is currently conceived within the project, including the design of the agents, the environment, the mechanism for the evolution of language and the peer-to-peer infrastructure on which the simulation runs.
Uploads
Papers by Bart Craenen