Decentralized discrete optimal quadratic automatic generation control of interconnected power systems based on the functional minimization method

Abstract

The distribution of electrical power over a large geographic area, the recent integration of distributed resources, the adoption of novel concepts like smart grids, and the digitisation of the power systems leads will undoubtedly increase the size of Multi-Area Interconnected Power Systems (MAIPS). Multi-area interconnected power systems are significantly complicated to control. Therefore, designing the Automatic Generation Control (AGC) is more complex and challenging. In order to control interconnected power systems, a robust AGC approach based on a Linear Quadratic Regulator (LQR) is being considered as a solution to these AGC problems. The LQR control approach provides wide stability margins, strong rejection of nonlinearities, reliability, simplicity, and good disturbance rejection. Despite the difficulty in determining the optimal state and controlling weighting matrices, the AGC design based on the LQR technique is not commonly employed. This study focuses on creating Automatic Generation control strategies based on Linear Quadratic Regulator and Functional Minimization Methods (FMM), and developing software programs and algorithms to address the AGC problems in MAIPS. In addition, such a complex system requires an analysis of the step load disturbances, modelling uncertainties, and nonlinearities of the MAIPS using decomposition and decentralisation approaches. According to the literature review, most AGC controllers are designed in continuous time and implemented in discrete time. The literature review findings show that the functional minimisation method (FMM) is suitable for constructing weighting matrices. The research work designed and implemented LQR-based AGC for the interconnected power systems in the discrete-time domain. To address the difficulties in AGC design, MAIPS has developed two AGC strategies, namely Discrete Centralized Optimal Quadratic Automatic Generation Control (COQAGC) and Discrete Decentralized Optimal Quadratic Automatic Generation Control (DOQAGC). DOQAGC is formulated for AGC problems in MAIPS based on optimal control theory and overlapping decomposition technique. The state and control weighting matrices are carefully selected for both developed control techniques using the functional minimisation process. At the same time, the overall system is divided into various subsystems via overlapping decomposition. The developed control methods and algorithms are tested in the MATLAB/SIMULINK environment with step load perturbations, modelling uncertainties, and generation rate limits (GRC), The selection of the states and control weighting matrices for both controllers is further streamlined and systematized by FMM. However, the DOQAGC's dynamic responses are greatly impacted by the fact that it disregards the effects of subsystem interconnections. The simulation findings show that DOQAGC is more reliable than COQAGC.