Implementation of cybersecure observers and control in microgrids; energy dynamics based approach

Abstract

Today's microgrids are modeled as dynamical systems with multiple physical components that interact. Microgrids are essentially small blocks that make up the larger energy grid and they are independent, meaning they can function separately and autonomously from the larger energy grid. These microgrids need to stabilize and produce the appropriate amount of power for the loads. In this thesis we adopt energy modeling for control and review rationale for claims that such an approach can achieve these goals. There are numerous types of control designs that can be used in these systems. A few examples include feedback linearizing control, conventional PID control, and energy control. This thesis discusses these types of control, and shows examples of each one used in a simulation. The examples are modeled and simulated using two main software tools: CAMPS (a MATLABbased Centralized Automated Modeling of Power Systems), and Simulink (an existing MATLAB tool). This thesis particularly emphasizes the implementation of these different control designs and their tradeoff. Each control design is used in an example in either CAMPS or Simulink, and the microgrids are probed at multiple points to compare results. Additionally, there are multiple ways to implement each control design; the tradeoff of the different methods are discussed. energy control is a novel technique used in microgrids - this thesis focuses on new implementation techniques of energy control using derivations from prior work in the field. Finally, an observer is introduced for supporting energy control so that control does not malfunction even when measurements are tampered. The first proof-of-concept simulation is provided to show this cyber-secure combination of energy control and energy observer for microgrids.