The Applied Study Of Fuzzy Control And Artificial Neural Network In The Voltage Stability

The power system is an essentially nonlinear dynamic system; its stability has been a major research topic. Voltage stability is an important research aspect in the stability of the power system. Voltage quality is one of the key indicators to measure power quality. To ensure the system's economy and reduce the loss of the power system, most of the power companies install the on load tap changer (OLTC) transformer and parallel compensation capacitor banks in the substation. By adjusting transformer tap running position and parallel compensation capacitor switching, it can ensure the system's voltage stability and economic operation at the same time.In the power system there are three primary factors affected the voltage's stability that is the parallel capacitors compensation, OLTC and the load inherent nonlinear characteristics respectively. This thesis presented a fuzzy control method based on the combination control of the compensator capacitor banks and the OLTC main transformer furnished in large-scale electric substation. By establishing the model with reactive power and voltage parameters based on the combination control of the compensator capacitor banks and the OLTC main transformer furnished in large-scale electric substation tested the fuzzy control system. In the control system, the change of voltage and power-factor were fuzzified as the input variable and the OLTC and the capacitor banks were controlled by the output signal which was calculated after fuzzy inference and defuzzifying. The result proves its stability and validity by simulation which combined with the realized substation parameters.By analysis all factors that impact voltage stability the OLTC negative effects and the load characteristics are two major reasons. This paper designs the OLTC negative effects early warning system and the load characteristics' early warning system using artificial neural network tool separately. In the design of the artificial neural network, by choosing a suitable intermediate layers and learning rate and momentum, the artificial neural network reduces production errors and improve the accuracy. By the design and training of artificial neural network, the two early warning systems can make a good prediction in the voltage stability, and improve the system's stability.In the end, the paper summarizes the main works that have been done, at the same time points out the existing problems and the focus of future research.