Communication-assisted adaptive protection can improve the speed and selectivity of the protection system. However, in the event, that communication is disrupted to the relays from the centralized adaptive protection system, predicting the local relay protection settings is a viable alternative. This work evaluates the potential for machine learning to overcome these challenges by using the Prophet algorithm programmed into each relay to individually predict the time-dial (TDS) and pickup current (IPICKUP) settings. A modified IEEE 123 feeder was used to generate the data needed to train and test the Prophet algorithm to individually predict the TDS and IPICKUP settings. The models were evaluated using the mean average percentage error (MAPE) and the root mean squared error (RMSE) as metrics. The results show that the algorithms could accurately predict IPICKUP setting with an average MAPE accuracy of 99.961%, and the TDS setting with a average MAPE accuracy of 94.32% which is sufficient for protection parameter prediction.
As conventional generation sources continue to be replaced with inverter-based resources, the traditional fixed overcurrent protection schemes used at the distribution level will no longer be valid. Adaptive protection will provide the ability to update the protection scheme in near real-time to ensure reliability and increase the resilience of the grid. However, knowing and detecting when to update protection parameters that are calculated with an adaptive protection algorithm to prevent unnecessarily communicating with relays still needs to be understood. The proposed method provides a sensitivity analysis to understand when it is necessary to issue new parameters to the relays. The results show that settings do not need to be issued at each available time step. Instead, the proposed sensitivity analysis method can be used to ensure that only the imperative protection parameters are communicated to the relay, allowing for more optimal utilization of the communications. The results show that the sensitivity analysis reduces the settings communicated to the devices by 93% over the year.
The integration of renewable and distributed energy resources to the electric power system is expected to increase, particularly at the distribution level. As a consequence, the grid will become more modular consisting of many interconnected microgrids. These microgrids will likely evolve from existing distribution feeders and hence be unbalanced in nature. As the world moves towards cleaner and distributed generation, microgrids that are 100% inverter sourced will become more commonplace. To increase resiliency and reliability, these microgrids will need to operate in both grid-connected and islanded modes. Protection and control of these microgrids needs to be studied in real-time to test and validate possible solutions with hardware-in-the-loop (HIL) and real communication delays. This paper describes the creation of a real-time microgrid test bed based on the IEEE 13-bus distribution system using the RTDS platform. The inverter models with grid-forming and grid-following control schemes are discussed. Results highlighting stable operation, power sharing, and fault response are shown.