RULE-BASED ANIMATION SYSTEM WITH GENETIC ALGORITHMS AS TEST-BED FOR GENERIC EVOLUTIONARY APPLICATIONS

The rule-based programming paradigm enhanced with genetic algorithms results in a versatile animation system for many applications in Artificial Life (AL) and creative design. We present some classes of applications of Genetic Algorithms (GA) realized with a parametric, timed, and stochastic L-system based embryogeny for phenotype generation. Many types of evolutionary applications can be simulated and visualized in a common phenotype space, the 3D virtual environment defined by L-systems. After the description of the rule-based embryogeny and the features of the GA, we present a standard example for function optimization. The second class of applications illustrates the system’s capability for evolutionary interactive design. The third example deals with evolutionary group animation. Interactive users feed populations of fish evolving according to their fitness. The examples show that a GA with the rule-based embryogeny can model many types of evolutionary applications coexisting in the same virtual 3D world. 1