DEVELOPMENT OF THE FUZZY PSS AUTOMATED DESIGN METHOD

Paper presents the method of the unified newo-flit y PSS automated design, based on hybrid technology ANTIS (a(haptive neuro-fir_zy inference system). Ibis technology includes the transformation of fuzzy system into the adaptive network which hers the property to train itself on a bride range of operating conditions. Using this technolotrl, it is possible to gel quality indexes, Which are similar to the results achieved notit- li ith the use of conventional Russian PSSX but having various gains of stabilifation channels in different operating conditions. Stabiliser suggested in this work has the .same structure as the traditional Russian PSS. The linear-quadr utic-Guus'siun (LOG) control scherne bras applied as a test control system to learn neuro-fuzzy PSS. The simplest "sin le machine infinite bits" system as ire ll as fourth-machine Ontario Hl'dro scheme were used for evaluation of effectiveness of the method proposed. The automatic control of recommended itself as an effective way of increasing small-disturbed and transient stability level of power systems. In connection with development of microprocessor engineering, the methods of adaptive regulation have become perspective, as the fixed parameters of all operating conditions and transients of large energy pools. In some cases it is fraught with disturbance of power system stability, which level can be decreased because of increase of' generator number with unfavourable parameters, use of underexcitation modes and complication of operating conditions.