An Expansion Planning Approach for Intelligent Grids with Speculative Parallelism

Abstract

Smart grid planning is a standard method used to reduce the net loss of distribution networks and improve the economic efficiency of power systems. As the distribution networks expand, the scale is getting more complex, leading to low efficiency and a long planning time when using traditional methods to implement new routes. To solve this problem, this paper proposed a new approach for intelligent grid expansion planning with speculative particle swarm optimization. First, the expansion planning model for the smart grid is established based on particle swarm optimization from the classical control system. Secondly, the model is parallelized with the speculative parallelism technique to overcome the influence of the internal control dependency and get rid of the original processing order. Finally, the parallel model is implemented on a distributed computing platform to improve the planning efficiency for complex intelligent grids significantly. Experiments show that the proposed approach can improve the efficiency by about 40% in an Apache Spark cluster consisting of 20 nodes compared with the conventional ones. Moreover, the proposed method can also fully utilize the distributed cluster’s computing resources.