• Forward Selection Component Analysis
• Ordinary Least Squares
• phasor measurement unit
• PMU placement problem

The climate change, mainly caused by the greenhouse production, is one of the biggest challenges of the 21st century, as confirmed by many global conferences. A possible solution is seen in a more intense use of the alternative sources, but their integration in the power system are strongly changing the concepts of production, transmission and distribution of energy. The system of the future can be summarised in two words: smart grid. To achieve this new generation of grid, the distributed automation are playing a key role for monitoring, assessing and control. The phasor measurement unit (PMU) is a device widely developed in the last twenty years that permits to measure some foundamental states of the power system. However, its cost is high and then, their placement is formulated as an optimization problem, where the number of measurement sites is minimised without the loss of the system observability. The thesis proposes a new approach based on the Forward Selection Component Analysis (FSCA) algorithm. The structure of the FSCA, its similarities with the well known Forward Selection Regression (FSR) and Principal Component Analysis (PCA) algorithms, is presented. The variables selected by FSCA are employed as regressors in a linear model defined with the Ordinary Least Squares (OLS) method and validated by a residual analysis. Finally, a comparison with other approaches is provided.