Optimization of power plant AVR parameters to improve transient voltage stability

Abstract

The electrical power supply structure of today's power systems is changing towards high shares of renewable power supply. This leads to a decreased number of conventional power plants. However, conventional power plants still play a major role when it comes to voltage and frequency control. Thus, the remaining power plants in the system together with the renewable power supply units have to carry out those control actions in the future. To cope with these ongoing changes in the power system, conventional power plant operators are working on increasing the flexibility of their power plants. One aspect of power plant modernization is exchanging the automatic voltage regulator (AVR) of the synchronous generator. Oftentimes an old lead-lag controller is replaced by modern proportional-integral-derivative (PID) control. This paper shows how to use particle swarm optimization to tune the parameters of such PID regulators to significantly improve transient voltage behavior. AVR optimization is not new to the scientific community. However, most approaches are using simple linear control circuits representing excitation system and synchronous generator as PT1-elements. This paper uses dynamic non-linear models to represent the excitation system and the power system components and, thus, gives a very practical insight. Optimization results are evaluated within a simple single generator system and a nine bus benchmark system.