Catálogo de publicaciones - libros

Research on qualitative spatial reasoning has produced a variety of calculi for reasoning about orientation or direction relations between point objects or line segments. Altough it is obvious that some calculi are more general than others, the exact relationships between the calculi have not been investigated thoroughly. We show that many well-known orientation calculi can be expressed in the more general calculus which allows to translate information from one calculus into another. In addition, we demonstrate that the mapping can be exploited to automate typically complex tasks like determining or verifying composition tables.

Classifying objects in computer vision, we are faced with a great many features one can use. This paper argues that diagrammatic representations help to comprehend properties of features. This is important for the purpose of deciding which features should be used for a given classification task. We introduce such a diagrammatic representation for a shape feature and show how it enables one to decide whether this feature helps to distinguish some categories given. Additionally, it shows that the proposed feature keeps up with other features falling into the same complexity class.

Acquiring knowledge about persons is a key functionality for humanoid robots. In a natural environment, the robot not only interacts with different people who he recognizes and who he knows. He will also have to interact with unknown persons, and by acquiring information about them, the robot can memorize these persons and provide extended personalized services. Today, researchers build systems to recognize a person’s face, voice and other features. Most of them depend on pre-collected data. We think that with the given technology it is about time to build a system that collects data autonomously and thus gets to know and learns to recognize persons completely on its own.

This paper describes the integration of different perceptual and dialog components and their individual functionality to build a robot that can contact persons, learns their names, and learns to recognize them in future encounters.

A rule selection scheme of evolutionary algorithm is proposed to design fuzzy path planner for shooting ability in robot soccer. The fuzzy logic is good for the system that works with ambiguous information. Evolutionary algorithm is employed to deal with difficulty and tediousness in deriving fuzzy control rules. Generic evolutionary algorithm, however, evaluate and select chromosomes which may include inferior genes, and generate solutions with uncertainty. To ameliorate this problem, we propose a recombinant rule selection method for gene level selection, which grades genes at the same position in the chromosomes and recombine new parent for next generation. The method was evaluated with application of designing the fuzzy path planner, where each fuzzy rule was encoded as a gene. Simulation and experimental results showed the effectiveness and the applicability of the proposed method.

In this paper, a framework for previous and new quasi-exact extensions of the A-algorithm is presented. In contrast to previous approaches, the new methods guarantee to expand every state at most once if guided by a so-called monotone heuristic. By that, they account more effectively for aspects of run time while still guaranteeing that the cost of the solution will not exceed the optimal cost by a certain factor. First a general upper bound for this factor is derived. This bound is where is (an upper bound on) the maximum depth of the search. Next, we look at specific instances of the algorithm class described by our framework. For one of the new methods a linear, i.e. much tighter upper bound is obtained: the cost of the solution will not exceed the optimal cost by a factor greater than 1 + . The parameter  ≥ 0 can be chosen by the user. Within a range of reasonable choices for , all new methods allow the user to trade off run time for solution quality. Besides that, the formal framework also serves for a comparison in terms of other algorithmic properties of interest, e.g. in terms of a necessary condition for state expansion.

The results of experiments targeting the minimization of (BDDs) demonstrate large reductions in run time when compared to the best known exact approach for BDD minimization and to previous relaxation methods. Moreover, the quality of the obtained solutions is often much better than the quality guaranteed by the theory.

Due to their simple and intuitive manner rules are often used for the implementation of intelligent systems. Besides general methods for the verification and validation of rule systems there exists only little research on the evaluation of their robustness with respect to faulty user inputs or partially incorrect rules. This paper introduces a gray box approach for testing the robustness of rule systems, thus including a preceding analysis of the utilized inputs and the application of background knowledge. The practicability of the approach is demonstrated by a case study.

Many real-world application domains that demand planning and scheduling support do not allow for a clear separation of these capabilities. Typically, an adequate mixture of both methodologies is required, since some aspects of the underlying planning problem imply consequences on the scheduling part and vice versa. Several integration efforts have been undertaken to couple planning and scheduling methods, most of them using separate planning and scheduling components which iteratively exchange partial solutions until both agree on a result.

This paper presents a framework that provides a uniform integration of hybrid planning –the combination of operator based partial order planning and abstraction based hierarchical task network planning– and a hierarchical scheduling approach. It is based on a proper formal account of refinement planning, which allows for the formal definition of hybrid planning, scheduling, and search strategies. In a first step, the scheduling functionality is used to produce plans that comply with time restrictions and resource bounds. We show how the resulting framework is thereby able to perform novel kinds of search strategies that opportunistically interleave what used to be separate planning and scheduling processes.

In this paper we describe a time management approach to distributed agent-based simulation. We propose a new time management policy by joining optimistic synchronization techniques and domain-specific knowledge based on agent communication protocols. With respect to our experimental results, we assume that our approach helps to prevent too optimistic event execution. Consequently, the probability of time consuming rollbacks is reduced in comparison to a pure time warp based solutions. The approach has been implemented as a synchronization service for the JADE agent platform . The paper concludes by the discussion of our experimental results and future improvements.

We present a novel approach to adaptive multi-agent programming, which is based on an integration of the agent programming language GTGolog with adaptive dynamic programming techniques. GTGolog combines explicit agent programming in Golog with multi-agent planning in stochastic games. A drawback of this framework, however, is that the transition probabilities and reward values of the domain must be known in advance and then cannot change anymore. But such data is often not available in advance and may also change over the time. The adaptive generalization of GTGolog in this paper is directed towards letting the agents themselves explore and adapt these data, which is more useful for realistic applications. We use high-level programs for generating both abstract states and optimal policies, which benefits from the deep integration between action theory and high-level programs in the Golog framework.

Several options are available to relate agent logics to computational agent systems. Among others, one can try to find useful executable fragments of an agent logic or use a model checking approach. In this paper, an alternative approach is explored based on the view that . Using the same starting point, one of the established agent logics, we ask instead if it is possible to construct a programming language for that agent logic. We show that the programming language and the agent logic are formally related by constructing a denotational semantics. As a result, the agent logic can be used as as a design tool to specify and verify the corresponding agent programs.

In particular, we construct an agent programming language that is formally related to the KARO agent logic. The KARO logic is an agent logic that builds on top of dynamic logic. The approach is based on mapping worlds in the modal semantics for KARO onto a state-based semantics. The state-based semantics can be used to define an operational semantics for KARO programs. In this way, we obtain a computationally grounded semantics for a significant part of the KARO logic, including the operators for knowledge or beliefs, motivational attitudes and belief revision actions of a rational KARO agent.