Uncertainty Modeling and Stochastic Control Design for Smart Grid with Distributed Renewables

In traditional electric grid planning, the uncertainty arises due to the random consumption patterns of households. However, in future smart-grids there is to be a second source of uncertainty due to the inherent intermittent natures of distributed renewable generations such as solar, wind and tidal resources at customer premises that would also be integrated to the electric grid. In this short paper, we propose a non-stationary Markov chain model for the time transient household load. A maximum likelihood estimator is also derived to estimate the time variant parameters of the Markov chain. We then develop a stochastic reference dynamics-based tracking scheme for the utility-maintained central power plant to ensure grid reliability in the presence of time-varying load demands and integrated renewable distributed generators (RDG´s). Optimal controller is derived for each tracking scheme and tracking performance simulation results are also presented.