Establishment of rule based fuzzy system with assigned transfer function

摘要

The generation of suitable linguistic values and rules for a fuzzy logic system having a number of inputs and outputs is based upon a transfer function in a neural network The transfer function is based upon increasing the single input parameters to form partial functions. The value range of the input parameters is defined within limiting intervals. The fuzzy sets are based upon triangular shaped membership functions. A rules base is formed to an optimising process to give a good approximation to the entered transfer function.