Gary B. Parker

Revitalizing computing education through free and open source software for humanity

The humanitarian focus of socially useful projects promises to motivate community-minded undergrads in and out of CS.

Punctuated anytime learning for hexapod gait generation

Punctuated anytime learning is presented as the solution for two problems: the use of anytime learning with an off-line learning module and the linking of the actual robot to its simulation during evolutionary robotics. Two methods of punctuated anytime learning, fitness biasing and the co-evolution of model parameters, are described and compared using the common task of gait generation for a hexapod robot with changing capabilities.

Co-evolving model parameters for anytime learning in evolutionary robotics

Abstract Evolutionary robotics is a research area that makes use of evolutionary computation (EC) to provide a means of learning in robots. In this paper, we discuss a new way of integrating the actual robot and its model during EC. This method, which involves the co-evolution of model parameters, is applied to the problem of learning gaits for hexapod robots. The form of EC used is the cyclic genetic algorithm (CGA). Tests done in simulation and on the robot show that the CGA operating on the co-evolving model of the robot can adapt to changes in the robot’s capabilities to provide a system of anytime learning.

The Evolution of Multi-Layer Neural Networks for the Control of Xpilot Agents

Learning controllers for the space combat game Xpilot is a difficult problem. Using evolutionary computation to evolve the weights for a neural network could create an effective/adaptive controller that does not require extensive programmer input. Previous attempts have been successful in that the controlled agents were transformed from aimless wanderers into interactive agents, but these methods have not resulted in controllers that are competitive with those learned using other methods. In this paper, we present a neural network learning method that uses a genetic algorithm to select the network inputs and node thresholds, along with connection weights, to evolve competitive Xpilot agents

Evolving Parameters for Xpilot Combat Agents

In this paper we present a new method for evolving autonomous agents that are competitive in the space combat game Xpilot. A genetic algorithm is used to evolve the parameters related to the sensitivity of the agent to input stimuli and the agents level of reaction to these stimuli. The resultant controllers are comparable to the best hand programmed artificial Xpilot bots, are competitive with human players, and display interesting behaviors that resemble human strategies.

Evolving gaits for hexapod robots using cyclic genetic algorithms

A major facet of multi-legged robot control is locomotion. Each leg must move in such a manner that it efficiently produces thrust and provides maximum support. The motion of all the legs must be coordinated so that they are working together to provide constant stability while propelling the robot forward. In this paper, we discuss the use of a cyclic genetic algorithm (CGA) to evolve control programs that produce gaits for actual hexapod robots. Tests done in simulation and verified on the actual robot show that the CGA successfully produces gaits for both fully capable and disabled robots.

Evolving cyclic control for a hexapod robot performing area coverage

For a robot to search an entire area, it must follow a path that allows the range of its sensors to cover all parts of the area. This problem is a subset of path planning called area coverage. Most work done in this type of path planning has concentrated on ways of dividing the area up to avoid obstacles while covering the area. This is an important step in the process, but often takes for granted the movement of the robot within clear areas. This is not a problem if the robot has sufficient calibration to ensure the accuracy of calculated turns or if it has accurate enough navigational devices to keep track of its location. However, simple legged robots usually lack both of these attributes. It is difficult to make turns that fit a specified arch and sufficient on board navigational devices are expensive and/or too large to carry. In this paper, we use cyclic genetic algorithms to learn the control cycles required to make an actual hexapod robot perform area coverage.

Using a Queue Genetic Algorithm to Evolve Xpilot Control Strategies on a Distributed System

In this paper, we describe a distributed learning system used to evolve a control program for an agent operating in the network game Xpilot. This system, which we refer to as a queue genetic algorithm, is a steady state genetic algorithm that uses stochastic selection and first-in-first-out replacement. We employ it to distribute fitness evaluations over a local network of dissimilar computers. The system made full use of our available computers while evolving successful controller solutions that were comparable to those evolved using a regular generational genetic algorithm.

Learning gaits for the Stiquito

It has been shown that the use of cyclic genetic algorithms can be an effective means of gait generation for hexapod robot simulations. They can, with only low-level primitives, produce reasonable gaits in minimal time. In addition, their output requires little in intermediate controller complexity as it is a sequence of these primitives, which can be fed directly into the robot. In this paper, we test the applicability of these algorithms on an actual robot. A model for simulation was produced based on the measured capabilities of the Stiquito robot. This model was trained with the CGA using five random initial populations; gaits quickly evolved for all five. Tests on the actual semi-autonomous robot showed that after 1000 generations gaits comparable to the best designed by human engineers were produced.

Evolving autonomous agent control in the Xpilot environment

Interactive combat games are useful as test-beds for learning systems employing evolutionary computation. Of particular value are games that can be modified to accommodate differing levels of complexity. In this paper, the authors presented the use of Xpilot as a learning environment that can be used to evolve primitive reactive behaviors, yet can be complex enough to require combat strategies and team cooperation. In addition, this environment was used with a genetic algorithm to learn the weights for an artificial neural network controller that provides both offensive and defensive reactive control for an autonomous agent.