Robust synchronous generator excitation regulator based on stabilizing feedback action

Abstract

Summary A new method for synchronous generator static excitation control is proposed in this paper, based on the PI controller and stabilizing feedback action. Compared with the existing solutions, this paper introduces improvements in the system's dynamic performance, robustness, and sensitivity in relation to noise, based on simplified control system structure modification. By introducing the stator voltage stabilizing feedback action, an improved level of dynamic performance is achieved, which also introduces an increased level of robustness. Furthermore, compared with sequential PID and lead–lag compensator based controllers proposed in related standards and technical papers, the system sensitivity in relation to the measurement noise is not compromised, since the differential control action is not included in the proposed controlling structure. The controller design procedure is performed by means of a frequency domain based technique, while its performance is verified by means of simulation and experimental test runs. Copyright © 2015 John Wiley & Sons, Ltd.

Control of synchronous machines set with microprocessorcontrolled excitation units in high-voltage tests station

“…There are also carried out tests involving gradual, relatively slow voltage increase during load current measurement. For this reason, the software of the generator's microprocessor system uses a simple PID type voltage controller with the ability to operate in PI mode [16][17][18][19]. Figure 4 in the range of 0-3637 V which corresponds to the phase-to-phase voltage of 0-6300 V (0-U N ).…”

Robust synchronous generator excitation based on novel feedforward control

“…However, to examine the robustness of controller 20 compared with controllers 21 and 22, the experiment was performed consisting of the SG operating in island run with significant inductive load, load X L = R L + jωL L (L L = 0.06 H and R L = 14.5 Ω). Namely, the island run experiment is chosen since it can successfully illustrate the impact of the SG By analysing the experimental traces given in Figure 8, it can be concluded that the significant variation of the load inductance X L value causes variation of the SG stator voltage rising time from 18 milliseconds up to 20 milliseconds for the novel controller (Equation 20), while for the cases of existing conventional controllers the rising time increases from 18 milliseconds up to 36 milliseconds for Equation 21 and up to 58 milliseconds for Equation 22. Consequently, it can be concluded that the new feedforward action (Equation 9) significantly improves excitation system robustness, when compared with solutions with no feedforward action, or with solutions with the feedforward action based on the load compensation.…”

“…In El‐Metwally, a neural network–based solution is presented, to optimally tune the chosen set of linear sequential excitation controllers. In other studies, different linear and nonlinear SG excitation controllers are proposed, which are robust in relation to the SG model parameter variations and variable modes of operation.…”

“…In El-Metwally, 18 a neural network-based solution is presented, to optimally tune the chosen set of linear sequential excitation controllers. In other studies, [19][20][21] different linear and nonlinear SG excitation controllers are proposed, which are robust in relation to the SG model parameter variations and variable modes of operation. In this paper, the survey is oriented mainly towards the conventional sequential excitation controllers, since they represent a majority of solutions used in the excitation systems, as a result of their simplicity of operation and commissioning. However, the main goal of this paper is to propose the feedforward control action that improves the robustness of any applied superimposed controller, including conventional linear controllers.…”