| This dissertation investigates local control techniques that are applied to individual components in a do power system. These local controls operate autonomously using locally obtained information, sensed at their respective components. Yet, they contribute to overall system stability without requiring a central controller or peer-to-peer communication network. In the event of a disturbance in the dc system, whether ephemeral or cataclysmic, autonomous local controls enable the system to self-heal in the sense that unaffected components still operate. Load priority ensures that if energy is limited, the most important loads have preferential access. Limited system knowledge, such as the overall health of the system or change in mission objective, can be used to fine tune the controller performance. However, the controllers still operate autonomously if access to this information is lost. These controls are ideal for high-reliability, advanced energy systems since they can perform system-level coordination and are fault-tolerant.; Applications considered are supply-side and demand-side management. On the supply side, droop control is examined as a form of autonomous local control that adds damping to stabilize the power system. On the demand side, control strategies are shown for both single bus and multibus systems. In a single bus system, dynamic load interruption is shown to be useful to prevent voltage collapse, when demand exceeds supply, and to be capable of automatic load restoration upon system stabilization. In a multibus system, autonomous local controls can ensure reliable system operation by reconfiguring how the load is supplied to ensure seamless power transfer during fault conditions or partial loss of generation. |
