Optimization, Control and Operation Strategies for Distributed Energy Systems

In this dissertation, the problem of optimal operation strategies and control for gas turbine based distributed energy systems (DES) is addressed using mathematical optimization methods and computational intelligence techniques. Detailed mathematical component models for the key components of the DES such as gas turbine power generator, heat exchanger, absorption chiller, photovoltaic and wind turbine, are developed to capture the important behavior of these components and their interactions with the overall DES. These models are used to develop integrated DES models for the purpose of optimization and control design. A multi-objective function considering the total system efficiency and operational cost is formulated for designing optimal operation strategies. Various optimization problems are formulated based on different DES configurations depending on if energy storage units or renewable energy are used, which results in nonlinear programming (NLP), mixed integer nonlinear programming (MINLP) and stochastic programming (SP) problems. A two-stage approach combining the improved particle swarm optimization (PSO) with the sequential quadratic programming (SQP) method is proposed to solve the resulting optimization problems. The simulation results are compared with those using traditional rule-based operation methods under various loads, such as summer, transition season and winter. It is found the proposed optimal strategy for the DES is capable of achieving an improved performance. It is illustrated that by applying renewable energy the operational cost will be reduced and the system efficiency is increased. A model predictive control with is employed for the real-time control implementation of the proposed optimal strategy. It is shown by applying the proposed two-stage method the DESs are able to follow the optimal strategies under all conditions