Large scale power systems are normally divided into control areas based on the principle of coherency. The coherent areas are interconnected through tie-lines which are used for contractual energy exchange between areas and provide inter-area support during abnormal operations. A mismatch in the real power balance affects primarily the system frequency, but leaves the bus voltage magnitude unaffected. The problem of real power balance in a power system is termed as Automatic Generation Control (AGC).Many investigations in the area of AGC of interconnected power system have been reported in the past. However, a few works have been reported in the area of AGC of an interconnected hydro-thermal system. The motivation behind this dissertation is to study the problem of AGC of interconnected hydro-thermal system. The dissertation deals with some aspects of AGC of interconnected hydro-thermal system considering AGC strategies based on conventional Area Control Error (ACE) in continuous mode. The hydro area is considered with electric governor and thermal area is considered with single reheat turbine.

Present work describes a comprehensive approach for optimizing the parameters of the conventional supplementary controllers for a two and three equal area hydro-thermal systems, with hydro area equipped with electric governor and thermal area equipped with single stage reheat turbine, using ISE technique. Efforts have also been made to optimize the electric governor parameters in the presence of GRC using ISE technique.

Investigations have been carried out to study the dynamic response with optimized conventional integral controllers for AGC of an interconnected hydro-thermal system .Proper selection of governor speed regulation parameter ‘R’ for primary control loop is extremely vital. No work seems to have been reported about proper selection of R for interconnected hydro-thermal system in continuous mode. Studies have been carried out to obtain possible large value of R, without jeopardizing system dynamic responses with conventional integral controllers. A larger value of R will make governor realization cheaper and simpler. Sensitivity analysis has been carried out to examine the robustness of the optimum integral controllers to wide changes in system parameters and loading condition from their nominal values. The effect of change in Tie-line reactance on system dynamics has also been studied.

The effect of change in Area Capacity when additional hydro plants are added to a two area hydro-thermal system is studied and dynamic responses obtained. The effect on the system dynamics due to fault occurring in the tie-line of an inter-connected hydro-thermal system at different instants and its clearance after few cycles have been studied and dynamic responses obtained.

Comprehensive study of AGC for a three area hydro-thermal system has been carried out. A Systematic approach using ISE technique has been used to obtain the integral controller gains and electrical governor parameters for hydro plant in the presence of GRC. Dynamic responses of three area hydro-thermal system with optimized values obtained for conventional integral controllers have been studied. Investigations have also been carried out to examine the sensitivity analysis of integral controllers and the effect of tie strength and tie switching on system dynamic response. Studies have also been carried out to examine the effect of area capacity on dynamic response for a three area hydro-thermal system.

In the present day environment the deregulation of power sector has started with the emergence of ‘independent power producers’ (IPP) that sell power to vertically integrated units (VIUs). The separation of generating companies (GENCOs), distribution companies (DISCOs) and Transmission Companies (TRANSCOs) have taken place. The traditional AGC has been modified for a hydro-thermal system taking into account of bilateral contract on the dynamics. The concept of DISCO participation matrix to simulate these bilateral contracts in the two area hydro-thermal system has been studied. The importance of area control error participation factor (APF) on system dynamics has been rigorously studied and discussed.

The application of “Artificial Neural Network (ANN)” on a real time basis is explored. The massive parallelism involved in neural nets reduces the computational time significantly and thereby giving us the optimized controller gains. ANN has been applied to AGC to obtain an adaptive controller providing optimum integral gains over the entire operating region of the system. Studies have been made to examine the effect of activation function, normalization, Number of hidden units, number of patterns for training the neural network.