A Comparative Study on PID Controller Tuning of Load Frequency Control in Three-Area Power System Using Soft Computing Technique

Abstract

A disparity between power demanded, and power generated by the system generators results in the frequency deviation of the power system. So, the aim of Automatic Load frequency control (ALFC) is to preserve the frequency to its nominal value, and its allowed range. Nowadays, power systems are interconnected and grouped by numerous control areas. Tie lines connect each control area, allowing electricity to be transferred between them. Thus, the second objective of ALFC is to keep the tie-lines power flow close to its planned value. So, an exact, and realistic LFC model is very essential. Many types of control tactics have been implemented by researchers so far to realize the objectives of LFC. In this thesis, essentially, three-area interconnected systems consisting of reheat, and hydro turbine with generation rate constraint (GRC) are presented. The parameters of the PID controllers gains (Kp, Ki , Kd) values are optimized by applying four different metaheuristics optimization algorithms, namely the Chimp optimization algorithm(ChOA), Novel quasi-harmony search algorithm (QOHS), Mothe- flame optimization algorithm(MFO), and Multi-verse optimization algorithm(MVO). The peak overshoot, undershoot, settling time, and steady-state error is the time domain specification used for measuring the controller's performance, and the suggested system employs integral time absolute error(ITAE) as an objective function. From the MATLAB simulation result, when 1% step load perturbation is applied to area 1, the QOHS PID tuning gave a better result in area -1, and area-2 while MFO-PID tuning gave a better result in area-3. But in reducing deviation in tie-line power error MFO-PID tuning gives better results in all areas. When 10%, 5%, and 1% SLP is applied to area 1, area 2, and area 3 simultaneously, ChOA based PID controller gave a better result in area 1, and area 3, while MFO-PID tuning gave a better result in area-2. Again in reducing deviation in tie-line power error ChOA-PID tuning gives better results in all areas. From an execution time efficiency point of view, ChOA shows superior performance. Robustness was also checked by varying system parameters 1% step load change in area 1 and found that the system is achieving a steady-state position within the permissible time frame. It is concluded that a single control design technique cannot satisfy all of the requirements, including robustness to system uncertainties, simple/systematic design, and nonlinearity handling.