Distributed parameter estimation for complex energy systems
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science.
MetadataShow full item record
With multiple energy sources, diverse energy demands, and heterogeneous socioeconomic factors, energy systems are becoming increasingly complex. Multifaceted components have non-linear dynamics and are interacting with each other as well as the environment. In this thesis, we model components in terms of their own internal dynamics and output variables at the interfaces with the neighboring components. We then propose to use a distributed estimation method for obtaining the parameters of the the component's internal model based on the measurements at its interfaces. We check whether theoretical conditions for distributed estimation approach are met and validate the results obtained. The estimated parameters of the system can then be used for advanced control purposes in the HVAC system. We also use the measurements at the terminals to model and verify the components in the energy-space which is a novel approach proposed by our group. The energy space approach reflects conservation of power and rate of change of reactive power. Both power and rate of change of generalized reactive power are obtained from measurements at the input and output ports of the components by measuring flows and efforts associated with their ports. A pair of flow and efforts is measured for electrical and gas ports, as well as for fluids. We show that the energy space model agrees with the conventional state space model with a high accuracy and that standard measurements available in a commercial HVAC can be used for calculating the interaction variables in the energy space model. A novel finding is that unless measurements of both flow and effort variables is used, the sub-model representing rate of change of reactive power can not be validated. This implies that commonly used models in engineering which assume constant effort variables may not be sufficiently accurate to support most efficient control of complex interconnected systems comprising multiple energy conversion processes.
Thesis: M. Eng., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, September, 2020Cataloged from student-submitted PDF of thesis.Includes bibliographical references (pages 81-83).
DepartmentMassachusetts Institute of Technology. Department of Electrical Engineering and Computer Science
Massachusetts Institute of Technology
Electrical Engineering and Computer Science.