Jimmy Pettersson

Structural Evolution of Central Pattern Generators for Bipedal Walking in 3D Simulation

Anthropomorphic walking for a simulated bipedal robot has been realized by means of artificial evolution of central pattern generator (CPG) networks. The approach has been investigated through full rigid-body dynamics simulations in 3D of a bipedal robot with 14 degrees of freedom. The half-center CPG model has been used as an oscillator unit, with interconnection paths between oscillators undergoing structural modifications using a genetic algorithm. In addition, the connection weights in a feedback network of predefined structure were evolved. Furthermore, a supporting structure was added to the robot in order to guide the evolutionary process towards natural, human-like gaits. Subsequently, this structure was removed, and the ability of the best evolved controller to generate a bipedal gait without the help of the supporting structure was verified. Stable, natural gait patterns were obtained, with a maximum walking speed of around 0.9 m/s.

Optimization of waypoint-guided potential field navigation using evolutionary algorithms

This paper describes a method for optimization of waypoint selection for potential field navigation in autonomous robots. In the method presented here, a genetic algorithm (GA) is used for optimizing the potential field. The chromosome of each individual encodes parametrizations for the potential field generated by waypoints, obstacles, and goals. The waypoints themselves are obtained through a Voronoi tessellation of the environment in which the robot is operating. It is demonstrated that the algorithm allows a robot to navigate safely and efficiently through spaces with many obstacles, even in cases where these are placed in a strongly unfavorable way. Furthermore, the results from simulations were implemented successfully in an actual Khepera robot. Using a slightly simplified navigation procedure, in which the robot comes to a standstill between successive steps in the navigation, the Khepera robot managed to navigate through one of the most difficult environments used in the simulations. Finally, the paper briefly describes a different implementation of potential field navigation, in the path planning adaptation submodule of a more advanced simulated mobile robot (VirBot).

Application of the utility function method for behavioral organization in a locomotion task

The generation of a complete robotic brain for locomotion based on the utility function (UF) method for behavioral organization is demonstrated. A simulated, single-legged hopping robot is considered, and a two-stage process is used for generating the robotic brain. First, individual behaviors are constructed through artificial evolution of recurrent neural networks (RNNs). Thereafter, a behavioral organizer is generated through evolutionary optimization of utility functions. Two systems are considered: a simplified model with trivial dynamics, as well as a model using full newtonian dynamics. In both cases, the UF method was able to generate an adequate behavioral organizer, which allowed the robot to perform its primary task of moving through an arena, while avoiding collisions with obstacles and keeping the batteries sufficiently charged. The results for the simplified model were better than those for the dynamical model, a fact that could be attributed to the poor performance of the individual behaviors (implemented as RNNs) during extended operation.

Behavioral Selection Using the Utility Function Method: A Case Study Involving a Simple Guard Robot

In this paper, the performance of the utility function method for behavioral organization is investigated in the framework of a simple guard robot. In order to achieve the best possible results, it was found that high-order polynomials should be used for the utility functions, even though the use of such polynomials, involving many terms, increases the running time needed for the evolutionary algorithm to find good solutions.

UFLIBRARY: A SIMULATION LIBRARY IMPLEMENTING THE UTILITY FUNCTION METHOD FOR BEHAVIORAL ORGANIZATION IN AUTONOMOUS ROBOTS

A simulation software package (UF Library) implementing the utility function (UF) method for behavior selection in autonomous robots, is introduced and described by means of an example involving a simple exploration robot equipped with a repertoire of five different behaviors. The UF Library (as indeed the UF method itself) is aimed at providing a rapid yet reliable and generally applicable procedure for generating behavior selection systems for autonomous robots, while at the same time minimizing the amount of hand-coding related to the activation of behaviors. It is demonstrated how the UF Library allows a user to rapidly implement individual behaviors and to set up and carry out simulations of a robot in its arena, in order to generate and optimize, by means of an evolutionary algorithm,the behavior selection system of the robot.

A General-purpose Transportation Robot An Outline of Work in Progress

An outline of a current joint project between Chalmers University of Technology (in Sweden) and several Japanese universities (Waseda University, Future University, and the University of Tsukuba) is presented. The aim of the project is to build a general-purpose transportation robot for use in hospitals, industries, and similar facilities. The project will provide a thorough test of the recently developed utility function method for behavior selection, which will be used for generating the decision-making system in the transportation robot. In this paper, an outline of the proposed transportation robot is given, along with a brief description of some of the challenges arising from this project. Furthermore, the utility function method is presented. Finally, the results obtained thus far are briefly discussed, and some directions for further work are provided

Behavioral selection in domestic assistance robots: A comparison of different methods for optimization of utility functions

In this paper, the performance of several evolutionary algorithms (EAs), involving different operators, is investigated in connection with the utility function (UF) method, a method for generating behavioral organization (selection) systems in autonomous robots. The standard UF method, which uses an ordinary genetic algorithm (GA) with fixed-length chromosomes is compared with modified evolutionary methods in which the chromosomes are allowed to vary in size. The results show that, contrary to expectations, the standard UF method performs at least as well as the modified methods, despite the fact that the latter have larger flexibility in exploring the space of possible utility functions. A tentative explanation of the results is given, by means of a simple, analytically tractable, behavioral organization problem.