Some distance measures for morphological diversification in generative evolutionary robotics

Abstract

Evolutionary robotics often involves optimization in large, complex search spaces, requiring good population diversity. Recently, measures to actively increase diversity or novelty have been employed in order to get sufficient exploration of the search space, either as the sole optimization objective or in combination with some performance measurement.When evolving morphology in addition to the control system, it can be difficult to construct a measure that sufficiently captures the qualitative differences between individuals. In this paper we investigate four diversity measures, applied in a set of evolutionary robotics experiments using an indirect encoding for evolving robot morphology. In the experiments we optimize forward locomotion capabilities of symmetrical legged robots in a physics simulation.Two distance measures in Cartesian phenotype feature spaces are compared with two methods operating in the space of possible morphology graphs. These measures are used for computing a diversity objective in a multi-objective evolutionary algorithm, and compared to a control case with no diversity objective.For the given task one of the distance measures shows a clear improvement over the control case in improving the main objectives, while others display better ability to diversify, underlining the difficulty of designing good, general measures of morphological diversity.

Overcoming Initial Convergence in Multi-objective Evolution of Robot Control and Morphology Using a Two-Phase Approach

“…With similar observations in our own work on co-evolving morphologies and control systems [10], in this paper we demonstrate a simple approach to reduce this effect. We demonstrate that it is possible to improve the solution quality after convergence of a morphology-controller evolutionary run by freezing the morphology, and allowing a second phase of continued evolution on the control system alone.…”

“…It has been pointed out that co-evolution of morphologies and controllers is difficult [9]. Methods which have been explored to tackle this challenge includes generative encodings [6,13,10], morphological diversity-enhanced algorithms [14,10,20], and more complex environments [15].…”

Memetic robot control evolution and adaption to reality

“…We have earlier conducted experiments on evolving a combination of legged robot morphologies and control systems in simulation and then constructing a selection of these in reality [26], [27], [28]. A recent investigation on the difficulty of co-evolving morphology and control in virtual creatures suggest that mutations in the morphology may cause large variations in the behavior of the controller [29], and this is in line with our own observations.…”

Morphological Evolution of Physical Robots through Model-Free Phenotype Development

“…To categorize the agents based on their morphology, a distance measure for the morphologies is needed [ 28 ]. Therefore, two quantitative measures of an agent’s morphology are introduced.…”