The multi level environmental neural network and hardware

Summary form only given, as follows. The principles of a neural network environmental model have been proposed. The principles are universal and can use different neural network architectures. Such a model is self-organizing, and it can operate in regimes with and without a teacher. It codes information about objects, their features, and the actions operating in an environment, and analyzes concrete situations. There are functions for making an action plan and for action control. The goal of the model is given from an external site. The model has more than sixteen active regimes. The multilevel environmental neural network is implemented; in software and hardware `NEUROLIT´ tools