Engineering Calculation And Algorithm Of Adaptation Of Parameters Of A Neuro-Fuzzy Controller

As part of the study, a control scheme with adaptation coefficients neuron-fuzzy regulator implemented. The area difference method used as training for network. It improved by adding rule base, which allows choosing optimal learning rate individual neurons neural network controller applied superstructure PID in process scheme. dynamic object can function different modes. This technological operates modes terms loading and temperature setpoints. Because experiments, power consumption amount time required maintaining same absorption process, using conventional neural-network evaluated. concluded that neuro-fuzzy reduced transient 19%.