Convex optimal control for plug and play microgrids

In this paper, we present a plug and play (PnP) approach for optimal control of commercial and military microgrid (MG). Operating today's MGs require professional knowledge of the system and calls for the operators to manually configure each of the components, especially in cases if intermittent renewable energy sources are present. As the technological advances allow low-cost microprocessors to be deployed on key components of the MG, we propose a new MG control approach based on the framework of Internet of things (IoT). This paper attempts to address the design of optimal controllers for a PnP MG. The developed theory and software are applicable to commercial and military MGs. To this end, we propose a novel IoT MG scheme, which is in between the traditional centralized and decentralized approaches. While the optimization is centralized, the data storage and process are distributed. The data sheet, together with runtime data are transmitted to the central controller for energy optimization. The contributions of this paper can be summarized as the followings. (1) We developed a charging slot organization (CSO) algorithm to transform the device data sheet information into the form of convex optimization. As a comparison, it is 591 times faster than Particle Swarm Optimization (PSO) in a realistic setup. (2) From the perspective of software systems, we established the XML schema to bridge the gap between PnP MGs and existing optimization solvers. The required information, such as characteristic curves or parameters, are downloaded to the microprocessors on the energy assets at the production phase; Then, the information is automatically aggregated to the MG controller during the run time. (3) Finally, we evaluate the algorithm with realistic weather and device data.