15-162 Cooperative Distributed Scheduling for Storage Devices in Micro Grids based on Gradient Evolution of KKT Multipliers and Consensus Networks

June 20, 2016

Are you interested in this project? Contact Us.

Authors

Abstract

Current commercial energy management systems are centralized, with a central node communicating with all other devices and accessing global and local information across units. This approach is vulnerable to single-point failures, jeopardizes the privacy of individual units, and will not scale efficiently as energy networks continue to expand and become more complex. This invention uses the concepts of dynamic KKT multipliers and consensus networks to replace the centralized system with a distributed network that is more robust, secure, and scalable.

 

With this system, individual nodes are equipped with controllers that possess local computation and prediction capabilities and can communicate with neighboring nodes. Using available local information, each controller performs computations and coordinates with its neighbors to predict global variables to optimize scheduling. No local information is disclosed during this process, so the privacy of individual nodes is preserved. Further, the absence of a central controller makes the system scalable and more resilient against single-point failures.

15162 Cooperative Distributed Demand Management

Do you agree to the terms and conditions regarding intellectual property contained in the Membership Bylaws signed by your company representative?