Nonlinear Modified PI Control of Multi-Module GCSCs in a Large Power System

This paper presents the design of a new control strategy for gate-controlled series compensators (GCSCs). GCSCs are new FACTS devices which can provide active power flow control on a transmission line. Proper placement of GCSCs in proximity to generators can also provide damping to system oscillations. This paper has investigated the effectiveness of multiple multi-module gate controlled series compensators (MMGCSCs) for large power systems. MMGCSCs can be less expensive devices with wide range of control of capacitive reactance in series with transmission lines. A nonlinear modified PI (NMPI) control is developed to provide power flow control and enhanced transient stability margin of the multi-machine power system. The NMPI is designed using a multi-layer neural network to approximate the blocking angle from the effective capacitive compensation provided by PI controller. A neural network with few neurons trained offline is used as an approximator /estimator for each MMGCSCs. This method has been shown effective for small and large disturbances on the IEEE 39 bus power system