A new approach for the distributed deployment of centralized algorithms in smart grids

Distributed algorithms are regularly used to deploy distributed paradigms in smart grids. Computations are executed without any centralized server even in a limited bandwidth network. However, there are still some drawbacks in the proof of convergence as well as a lack of programming abstraction. In this paper, we propose a new approach to the distributed paradigm by using distributed programming in a data manipulation language called Smartlog. A distributed programming methodology will automatically divide a centralized Smartlog program into multiple rules while ensuring the correct transformation of the centralized computations. An application of voltage control in a distribution grid with high penetration of PV systems is used as an illustration of this approach. Two implementations are compared: centralized programming and distributed programming. Both are deployed in a real-time simulation with OPAL-RT and a network of Raspberry Pis. The response time of each implementation is analyzed to evaluate their respective performance, showing that the Smartlog language is particularly appropriate to smart grids in terms of compacity and simplicity.