Design And Optimization Of Load Frequency Control Systems

Current power grids are composed of multiple interconnected power systems. With the development of rapid booming electric power market in China, the interconnection among power subsystems is increasing, The frequency and voltage of the power system must be ensured stable. The frequency variation depends on the active power, and the voltage variation is associated with the reactive power. Therefore, the stability of power system control problem can be decomposed into two independent control problems. One is the active power and frequency control which is called the load frequency control and the other is the reactive power and voltage control. This thesis studies the load frequency control problem and the main contents are as follows:First, since the current load frequency control strategies are complex, and the orders of the controllers are high, it is difficult to apply them in practice. This thesis proposes a PID controller design method via dynamic matrix control (DMC), and applies it to the load frequency control problem. DMC does not need a detailed mathematical model of the controlled plant thus is simple and can achieve good performance. However, receding horizon DMC demands large computation efforts and is not easy to implement in a conventional distributed control system (DCS). So this theis approximates a DMC to a PID controller utilizing its closed-loop form. Simulation results show that the proposed method can achieve good performance, with better load disturbance rejection and robustness.Secondly, to overcome the inaccuracy in power system modeling and to improve the disturbance rejection performance, a robust control strategy (Tornambe controller (TC)) was proposed using active compensation, This thesis applies TC controller in a four-area load frequency control system, and compares it with a decentralized state feedback method. Simulation results show that the proposed controller is effective and ensures favorable control effect even for very complex power systems.In order to optimize the load frequency control system, this thesis adopts an adaptive differential evolution algorithm. The algorithm is modified with the basic idea that the global search ability should be strengthened at the beginning and the local search ability should be strengthened in the later in order to keep the population diversity and improve the convergence speed and accuracy of the algorithm. Simulation results for a two-area power system show that the optimized PI controller with the proposed algorithm has better performance than the PI controller optimized by traditional optimization algorithms, that is, that the system has a good dynamic performance as well as a good robust performance.Finally, this paper adopts linear active disturbance rejection control technology (LADRC) for load frequency control systems. LADRC can treat various uncertain factors as a disturbance, and then through an extended state observer estimate and compensate it, therefore this method can obtain good decoupling effect, and enhanced robustness. Since the active disturbance rejection control needs to set a large number of parameters, the differential evolution (DE) algorithm is used to tune those parameters. Simulation results show that the optimized LADRC can achieve good control performance.In summary, the proposed design and optimization methods for the load frequency control problems improve significantly the performance of the LFC systems, and the final controllers are easy to implement in practice.