ABSTRACT
Evolution by variation and selection is nature's foundational method of problem solving. The evolutionary selection circuits model proposes the successive modification of enzymatic neurons through a process of variation and selection as a means of developing new information-processing functions in the brain. The structure of the Learn II simulation system is depicted schematically. Modules VARIATION, SELECTION, and PROPAGATION incorporate the basic functions of the selection circuits model. Modules TASK, INTERFACE, and NEURON are accessed by the SELECTION module. Analyzed the adaptive and computational capabilities of systems based on continuous enzymatic neurons using a robot navigation task. In this type of task, a robot moving on a plane must reach a fixed target in the presence of randomly directed mechanical forces. Each robot is controlled by a network of four continuous enzymatic neurons. When the robots face randomly directed mechanical forces at each step of the execution of the task, the uncertainty of the environment increases considerably.
