Bonifacio Castaño

Statistical distribution of generation-to-success in GP: application to model accumulated success probability

Many different metrics have been defined in Genetic Programming. Depending on the experiment requirements and objectives, a collection of measures are selected in order to achieve an understanding of the algorithm behaviour. One of the most common metrics is the accumulated success probability, which evaluates the probability of an algorithm to achieve a solution in a certain generation. We propose a model of accumulated success probability composed by two parts, a binomial distribution that models the total number of success, and a lognormal approximation to the generation-to-success, that models the variation of the success probability with the generation.

A ZigBee and RFID hybrid system for people monitoring and helping inside large buildings

This article refers to a system that combines ZigBee and RFID technologies for monitoring people behavior when moving inside a building. The RFID part of our system consists of a set of detector placed at strategic points. To accomplish this detection goal the visitors of the building have to be provided with a RFID tag either active or passive. The ZigBee wireless subsystem will send the information generated by the detectors to the main computer. The movements of all and every one of the attendees will be followed on-line and recorded in a database. Filed information will be useful to know people activity and performance afterward. The system also includes an artificial intelligent program (a planner) that would be used to direct the movements of visitors when necessary, for instance, in case of fast evacuation of the building for fire or earthquake.

3Dana: A path planning algorithm for surface robotics

Abstract Autonomous navigation is a research topic that has received considerable attention in robotics. Generally, it is a two step process: (i) generate a global route to the goal and (ii) local motion of the robot along the route. The focus of this paper is on the first part of the process. Some common techniques used are based on heuristic search algorithms that obtain (sub)optimal paths by usually exploiting a rather simplistic terrain representation. Then, the paths generated hardly take into account relevant terrain features, which leads to potentially unsafe paths in realistic environments. This paper presents two contributions: a mathematical formulation for any DTM that can be used by heuristic search algorithms, and a path planning algorithm that generates candidate paths that are safer than the ones obtained by previous approaches. This algorithm, called 3Dana, considers different parameters to maximize the path quality: the maximum slope allowed by the robot and the heading changes during the path. These constraints allow discarding infeasible paths while minimizing the heading changes. To demonstrate the effectiveness of the algorithm proposed, we present results for different scenarios, which include an evaluation of the algorithm in real Mars DTMs.

An empirical study on the accuracy of computational effort in Genetic Programming

Some commonly used performance measures in Genetic Programming are those defined by John Koza in his first book. These measures, mainly computational effort and number of individuals to be processed, estimate the performance of the algorithm as well as the difficulty of a problem. Although Kozas performance measures have been widely used in the literature, their behaviour is not well known. In this paper we study the accuracy of these measures and advance in the understanding of the factors that influence them. In order to achieve this goal, we report an empirical study that attempts to systematically measure the effects of two variability sources in the estimation of the number of individuals to be processed and the computational effort. The results obtained in those experiments suggests that these measures, in common experimental setups, and under certain circumstances, might have a high relative error.

On the data structure straight-line program and its implementation in symbolic computation

In this paper we describe a recent computer implementation (the PASCAL program TERA) of a well known Computer Algebra algorithm. The particularity of this implementation consists in the fact that it is based on a special abstract data type, namely that of a directed acyclic graph (DAG) which is of seldom use in Computer Algebra packages. This data type is particularly adapted to the algorithmic problem which we are considering in this paper: the computation of the of two multivariate polynomials. This task is solved by an algorithmic approach based on linear recurring sequences (see [F.R. Gantmacher, The Theory of Matrices, vol. 1/2, Chelsea, New York, 1959; R. Sendra, J. Llovet, Journal of Symbolic Computation 13 (1992) 25–39; J. Llovet, R. Sendra, J.A. Jaen, R. Martinez, Computer Science, 1992, pp. 159–165; R. Martinez, Procedimientos de Recurrencia Lineal en Algebra Computacional, PhD Thesis, Depto. de Matematicas, Universidad de Alcala de Henares, Espana, 1992; J. Llovet, R. Martinez, J.A. Jaen, Journal of Computational and Applied Mathematics 49 (1993) 145–152]). An experimental study shows that the time and memory space performance of the TERA program improves significantly upon that of traditional Computer Algebra Systems (MAPLE and MAGMA in our case).

Optimization of Trading Rules for the Spanish Stock Market by Genetic Programming

This paper deals with the development of a method for generating input and output signals in the Spanish stock market. It is based on the application of set of simple trading rules optimized by genetic programming. To this aim we use the HeuristicLab software. To evaluate the performance of our method we make a comparison with other traditional methods such as Buy & Hold and Simple Moving Averages Crossover. We study three different market scenarios: bull market, bear market and sideways market. Empirical test series show that market global behavior has a great influence on the results of each method and that strategies based on genetic programming perform best in the sideways market.

Integrating a PDDL-based planner and a PLEXIL-executor into the ptinto robot

Ptinto is a prototype of a hexapod robot for exploration tasks in rocky and cumbersome areas. The main objective of this prototype is to design and test a complex kinematic control system, including both new hardware and software technologies. In this paper we describe the autonomous architecture that we have developed for the control of the system. We use a deliberator that must be able to make a safe trajectory between two or more points avoiding obstacles. When a trajectory has been created, the executive takes this plan to control Ptinto via the hardware abstraction layer. This is a classical 3T architecture implementation with two general purpose systems: a PDDL planner as the deliberator, and a PLEXIL executor.

PIPSS*: A System based on Temporal Estimates

AI planning and scheduling are two closely related areas. Planning provides a set of actions that achieves a set of goals, and scheduling assigns time and resources to the actions. Currently, most of the real world problems require the use of shared and limited resources with time constraints when planning. Then, systems that can integrate planning and scheduling techniques to deal with this kind of problems are needed.

Simulation of the Hexapod Robot PTinto Walking on Irregular Surfaces

In this article we present our recent work on the simulation of a hexapod robot called PTinto. This robot is a prototype that has been designed to test the six legs locomotion system in the surrounding areas of the Tinto River (Huelva-Spain), which scientists have considered that could potentially have many similarities to the Mars surface. This kind of robots represents a great advance in the planetary research that overcomes the performance of the usual rover when operating in rocky and cumbersome areas. We are developing some of the software PTinto requires to autonomously control it and, in particular, the program that simulates its movements on any surface. Up to now we have a graphic computer program, developed in MATLAB, that represents the robot and allow us to check its movements in a lot of detail. We present in this paper all the elements and resources we use for the simulation and control of PTinto walking on irregular surfaces. Thanks to this simulation program, we could test different autonomous control strategies for the PTinto robot. (Received in October 2013, accepted in July 2014. This paper was with the authors 3 months for 1 revision.)

3Dana: Path Planning on 3D Surfaces

An important issue when planning the tasks that a mobile robot has to reach is the path that it has to follow. In that sense, classical path planning algorithms focus on minimizing the total distance, generally assuming a flat terrain. Newer approaches also include traversability cost maps to define the terrain characteristics. However, this approach may generate unsafe paths in realistic environments as the terrain relief is lost in the discretisation. In this paper we will focus on the path planning problem when dealing with a Digital Terrain Model (DTM). Over such DTM we have developed 3Dana, an any-angle path planning algorithm. The objective is to obtain candidate paths that may be longer than the ones obtained with classical algorithms, but safer. Also, in 3Dana we can consider other parameters to maximize the path optimality: the maximum slope allowed by the robot and the heading changes during the path. These constraints allow discarding infeasible paths, while minimizing the heading changes of the robot. To demonstrate the effectiveness of the algorithm proposed, we present the results for the paths obtained for real Mars DTMs.