Power Systems State Estimation Results - COCOON Project

Description of the PV Plant

The analysis is based on a PV plant layout consisting of two Medium Voltage (MV) feeders, each equipped with 3 inverters. The inverters are connected to the MV feeder through transformers, and the plant evacuates energy to the main grid via an electrical substation.

The following files in the root directory define the system's topology and monitoring setup:

• pv_2_3.svg: A graphical representation of the PV plant. Measurements are assumed to be available at all points marked with red squares in this diagram.

• pv_2_3.json: A data file containing the technical parameters and connectivity of the network.

• meas.zip: A compressed archive containing the measurements considered for the state estimation analysis.

Repository Structure

The code and results are organized into multiple folders, each representing a specific test case or scenario. Each folder follows a consistent internal structure:

/test_case_name/
│
├── data/      # JSON files containing results from simulation tests
├── figs/      # Graphical representations (plots, charts, etc.) of the results
└── README.md  # Interpretation and explanation of the specific test case

Folder Details:

• data/: Contains .json files with the output data generated from the simulations. These files can be used for further analysis or validation.

• figs/: Includes visualizations of the simulation results (PDF format). These plots help in understanding trends, anomalies, and overall system behavior.

• README.md: Each test case includes its own documentation offering key findings and relevant notes for understanding the results.

Analysis Methodology

All test cases are analyzed from two perspectives to evaluate system resilience:

1. Detection Phase: Focuses on determining whether a cyberattack has occurred. This involves evaluating the system's ability to distinguish between normal operation and compromised scenarios.

2. Identification Phase: Aims to identify which specific measurement(s) have been manipulated by the attacker. This allows for a deeper understanding of the attack vector and its impact on the system.

For both aspects, relevant metrics are provided to quantify the performance of the detection and identification processes, as well as their effects on the accuracy of the state estimation.

If further details are required or any questions need clarification, please contact Álvaro Rodríguez at arnozal@us.es.