Complex neural networks constructed by different kinds of neuron network models and neurocontrol

A novel concept of a neuron network model is submitted and a new method for constructing a type of complex neural network using different kinds of neuron network models is offered. The method is derived from the discovery of neurobiology and considered from the angles of reliability or robustness, topology, bionics and synagetics. The architecture and the corresponding property of this kind of neural network are in conformity with that of the human nervous system. This method can also be used in neurocontrol systems.<>