Decentralised Model Predictive Control with Integral Sliding Mode for Frequency Regulation in Power Grids

This paper proposes a decentralised Integral Sliding Mode (ISM) scheme coupled with a Model Predictive Control (MPC) architecture to control the frequency in power systems partitioned into control areas and characterised by unknown disturbances and unmodelled dynamics. The frequency and the total electrical active power exchanged are both locally measured in each area by Phasor Measurement Units (PMUs). A robust ISM control scheme with chattering alleviation is designed to compensate bounded uncertainties. Thanks to the achieved compensation, it is shown in the paper that a standard MPC control law for linear systems can be used in order to both regulate the frequency deviation and to minimise a quadratic cost function depending also on the cost of the generation. In addition, a method to locally and dynamically reconstruct the unknown disturbances exploiting ISM is designed. Numerical simulations are reported to verify and validate the formulated methodology.