Abstract
We have studied integrated electric utility networks in terms of long-term electric utility resource planning, short-term environmental economic load dispatch, and the optimal deployment of distributed generation sources within the electric network. In order to conduct these analyses, the development of a continuous-discrete modular simulation and optimization framework and a particle filtering based multi-objective optimization framework has been done. The continuous-discrete modular simulation includes the integration of discrete decisions and events with continuous processes within the same simulation environment and the modular integration of different resources into the simulation. The multi-objective optimization framework includes the development of a sequential importance sampling mechanism for multi-objective optimization, leveraging the information contained within the non-dominated set of solutions to increase the thoroughness of the generated solution and ensure that the algorithm does not converge to local optimums. The evaluation of energy capacity plans include the use of conventional energy generation sources, renewable energy generation sources, energy storage alternatives and the evaluation of environmental policies (in terms of their effect on energy capacity plans and their effectiveness in incentivizing environmentally friendly energy generation sources). The optimization of the environmental load dispatch is achieved leveraging Bayesian particle filtering to evaluate the state of the load dispatch at each generation unit, and the Newton-Raphson method to ensure that power balance constraints are met. The optimization of different levels of penetration of distributed generation has been evaluated in terms of minimal cost, power loss and environmental impact.