ABSTRACT--This paper presents an Automatic Generation Control (AGC) system in a two-area power system with a fractional order PID (FOPID) controller. In the proposed method, a newly developed harmony search algorithm is implemented to solve the non-convex non-linear problem of AGC. The fractional order PID controller will improve the system stability and improve the robustness of the controlled system. The proposed FOPID controller also has a smaller number of parameters to tune than its existing counterparts. In order to show efficiency of the proposed approach, the response of HSA-Optimized system to disturbances is compared with the results of GA-Optimized system. The cost function of the proposed problem is considered based on Integral of Time Multiplied Square Error (ITSE). In order to improve the system performance, frequency bias factor (FBF) is modeled in the problem and the proposed method is implemented in a non-linear interconnected two area power system. Simulation results verified that employing the proposed approach in optimization process of control system would guarantee a better system performance comparing to other methods.
Keywords-Integral of Time Multiplied Square Error;Harmony Search Algorithm; two-area nonlinear power system; Load-Frequency Control.
I.INTRODUCTIONDay to day the load demand is growing at very high rate. Due to surplus demand of electricity with increase in population and industrialization, the power generation with conventional sources was a great challenge to the generation stations to fulfill the demands. So, modern human societies have increased the significance of optimum Automatic Generation Control (AGC). By overlooking a well-designed AGC, the electric power system will face a lot of complicated difficulties. In other words, in an optimal operation of power system, it is indispensable that frequency deviations must maintain in allowable limits by a welldesigned AGC [1].Many papers have been published on the AGC problem considering a linear model. In [2], AGC is employed by utilizing linear feedback. Optimal control and neural network-based controller are some other methods employed in the AGC of a linear power system [3]. Additionally, some extra papers have been published on controlling non-linear model of the governor, generator and power system [5][6][7]. These methods are significantly time-inefficient in the AGC of power system.Regardless of all the developments in control methods, fractional order Proportional-Integral-Derivative (FOPID) controllers are still the most common type of controllers in industries. Therefore, it is sensible that a lot of papers have been published in these years on adjusting FOPID controller parameters using optimization algorithms [8][9][10]. There are different criteria employed in these papers to evaluate effectiveness of the enhanced controllers. Some papers have working an Integral of Absolute Error of system response to a step demand change [11][12], and some others have used an Integral of Square Error to the same disturbance ...