Wide Area Damping Control Based On Fuzzy-Neural Network

At present, the development of the electric power industry of China is in animportant period of inter-regional power grid interconnection, the national networkand the realization of the optimal allocation of resources. With the advent ofinterconnected system, power grid dynamic stability problem is becoming more andmore complete, one of the main challenges facing large-scale power grid is theinterval weak damping. Weak damping inter-area low frequency oscillation restrictsthe stability of the regional interconnected power grid transmission capacity andinterconnected system stability. So the design of good damping controller canimprove the system damping effectively, which is to damp the interval low frequencyoscillation and to improve the security and stability of the interconnected system, andit is of great significance to promote the development of electric power market.In view of the traditional damping controller which uses the local units signalsconstituting a feedback control system and can t inhibit inter-area low frequencyoscillation problem well. In this paper, the control system is built based on the widearea measurement theory. The paper takes the control system of the two district fourgenerators as an example, constructs the model in the PSASP environment for smallsignal stability analysis and finds out the main low frequency oscillation mode in thesystem, to analyze and select the leading units in the interval low frequencyoscillation of system, and finally uses analysis results to determine the selection ofcontrol installation location and the feedback information of wide area dampingcontroller.Considering the off-line designed and fixed parameters damping controllerswhich are based on detailed system modeling, the damping control is poor undercertain conditions. Combining the superiority of fuzzy controller avoiding the detailedsystem modeling requirements with advantage of neural network s robustness andartificial self-learning technique, self-derivation ability, this paper is applying fuzzyneural network technology in wide area supplementary damping controller design todamp the low frequency oscillation of interconnected power system. The input signalsof the damping controller can reflect the interval information of low frequencyoscillation system, and controller structure is the organic combination of neuralnetwork and fuzzy control, the neural network (NN) nonlinear mappingcharacteristics is so strong to memorize the fuzzy control decision rules. Finally the signal is added to the generator excitation system. Wide area damp controller avoidsthe previous wide area damping controller design problem of depending on theinterconnected systems model, and can adapt to the nonlinear characteristics of theinterconnected power systems damping control. Finally, the effectiveness of wide areadamping controller suppressing the low-frequency oscillation is verified in the twoarea four machine system in MATLAB/SIMULINK, and the damping controller canimprove the system stability effectively.