HIERARCHICAL AND DISTRIBUTED OPTIMAL CONTROL OF INTEGRATED ENERGY SYSTEMS

For decades the main goal of the power system was to deliver power from generation plants to consumers through an extensive high-voltage transmission system and a medium to low voltage distribution system. As such the existing power system was designed in a centralized tree like fashion in order to connect a relatively small number of large plants to a large number of consumers. In addition, the gas infrastructure is developed separately, with an infrastructure consisting of high, medium and low pressure pipelines, more or less similar to the power grid structure. Both grids are regulated separately, even though they are coupled. So far, the analysis and design of new smart energy systems in the energy infrastructure have been mainly focused on only one particular grid such as the power grid, gas grid or heat grid. In particular, the power grid has received by far the most attention. Much of the current research evolves towards embedding of renewables such as photovoltaic cells and wind power into the power grid, where it is important to take into account the cloud and wind conditions, which are less predictable than the power production of a classical power plant. In addition, there is a fine meshed gas network in the Netherlands, where embedding of biogases and/or non-natural gases is developed. Currently, this network is coupled to the power grid on a limited basis, i.e., some Combined Heat Power (CHP) systems that produce both heat and power and run on gas are embedded. Furthermore, some larger power plants operate on gas. In the future, the embedding of many more micro-CHP systems in households is foreseen, and storage of surplus electricity in the form of power-to-gas (hydrogen) is also foreseen. Within this research proposal we aim at developing new/tailored methods that integrate hierarchical and distributed optimal control methods for large-scale networked systems necessary for the embedding of new energy systems in the coupled power and gas grids. By a large-scale networked system, we mean a collection of interconnected heterogeneous dynamical systems (agents) sharing/collecting information. Thus, the research of this proposal is concerned with designing integrated distributed controllers (at the level of agents) based on the available limited local (asynchronous) information exchange in order to achieve a global (at the level of the entire network) hierarchical optimality criterion. To the best of our knowledge, there do not yet exist control schemes with strong theoretical foundations to ensure stability and optimality for such complex structures.