Voltage Stability Assessment Based on BP Neural Network

An assessment approach on power system voltage stability is provided using Back Propagation (BP) Neural Network, which takes the Voltage Collapse Proximity Indicator (VCPI) as assessment index. The key feature of the method is to establish static and dynamic assessment models on voltage stability. The training results of the static models based on load flow calculation can reflect the nonlinear mapping relationship correctly between power flows and voltages on load bus with given load increasing mode; Based on integrated load model, the dynamic model uses two three-layer BP networks to make classification and prediction on system, respectively. With two instances of WSCC-9 and 3 generator-12 bus power system, it is verified that the method is effective to voltage stability assessment on power system.