Risk Management and Combinatorial Optimization for Large-Scale Demand Response and Renewable Energy Integration

摘要: To decarbonize the electric power grid, there have been increased efforts to utilize clean renewable energy sources, as well demand-side resources such loads. This utilization is challenging because of uncertain generation and inelastic demand. Furthermore, interdependencies between system states networks or interconnected loads complicate several decision-making problems. Growing interactions systems human agents with advances in sensing, computing communication technologies also increase need for personalized operations.In this dissertation, we present three control optimization tools help overcome these challenges improve sustainability systems. The first tool a new dynamic contract approach direct load that can manage financial risks utilities customers, where are generated by generation. key feature proposed method its risk-limiting capability, which achieved formulating design problem mean-variance constrained risk-sensitive control. globally optimal contract, develop programming solution based on novel dynamical track limit risks. performance framework demonstrated using data from Electricity Reliability Council Texas. second developed combinatorial under interdependencies, inherent networks. For problems, be formulated systems, linear approximation scalable has provable suboptimality bound. algorithm illustrated ON/OFF supermarket refrigeration last seeks provide mechanism loads, play an important role management. We integrate Gaussian progress regression into model predictive learn customer's preference online automatically customize controller directly affect comfort. Finally, discuss future research directions operation sustainable cyber-physical including unified risk management electricity markets, selective resilient grids, contract-based modular infrastructure