Implementation of ANN based tap-changer control in power transmission and distribution system

A tap changer control operates to connect appropriate tap position of winding in power transformers to maintain correct voltage level in the power transmission and distribution system. A recently developed Artificial Neural Network (ANN) based tap changer control showssignificant improvement in the accuracy of tap-changing decisions [1,2]. This improved tap-changing decision and operational flexibility of this new technique over the existing tap changing controls make it attractive for deployment in practical power system network. This paper deals with the implementation of ANN based tapchanger control practically, using special purpose digital hardware as built-in semiconductor chip or software simulation in conventional computers. Two strategies are suggested for its implementation as asoftware module in the paper. One is to integrate it with the supervisory system in a substation control room operating in a LAN environment. In this configuration, the parallel transformers can be controlled locally. The other is to integrate it into the SCADA system, which allows the transformers to be monitored and controlledremotely over a wide area of power network. The implementation of ANN based tap changer control needs interfacing between the power system and the control circuitry, which is also presented in the paper.