The demo 8 started in the second year of the project, and the main activities performed during this period include:
- Site demonstrator visit and identification of necessary data to be provided by plant owner and network operator.
- Preliminary works on the steady-state and dynamic-state models, that were able to be run over PSS/E.
- Migration tasks of those models from PSS/E to ePhasorSIM have already started.
- State of the art review of research studies in predicting the motion of electromechanical modes in a given system, and development of linear and non-linear prediction models for predicting the modes of the system.
- Research of the state of the art in both exciter-based and governor-based PSS.
- The research showed that exciter-based PSS would help to kill local oscillations, while governor-based PSS would be useful to kill inter-area oscillations.
- Feedback of trials evaluation shows that the steady-state and dynamic-state models were loaded in PSS/E and it was possible to start the simulations with a stable system. Analysis comparing Deep Neural Networks and Linear Regression to predict the models of the system have started, with promising results, although some tuning is still required.
Latest works on this area have covered the analysis and experimentation on a solution consisting in a combination of grey-box and black-box models to update a dynamic PSS, to more effectively damp low-frequency oscillations than with traditional PSS. Progress has also been made on a Phasor Power Oscillation Damping (POD) PSS system, finding that with accurate system knowledge, the Phasor-POD PSS converges faster than traditional PSS systems. Besides, a self-tuning based approach is being designed for the PSS gain selection. Regarding more experimental tasks, a new AVR with a built-in PSS suitable for the regulation strategies that will be implemented, has been already delivered, assembled and wired at COGEN’s plant.
Regarding papers, a paper has been sent to CIGRE conference in Vienna: http://cigrevienna2020.at/
The extended abstract was accepted and we are waiting for the full paper acceptance. Find it enclosed.