A Parallel Algorithm of Optimal Power Flow on Hadoop Platform

Application of smart grid leads to significant increase in scale and data of the power systems, bringing new challenges to the calculation of optimal power flow. However, the existing parallel algorithms, such as MPI-based solutions, suffer from high computational complexity. In this paper, we propose a parallel algorithm of optimal power flow based on Map-Reduce framework. More concretely, the node reordering in our algorithm can greatly accelerate solution speed of the linear equations meanwhile fit well with Map-Reduce programming specifications. Moreover, we determine the appropriate formats for input, intermediate and output data sets and partition the algorithm into separate map/reduce tasks. This facilitates our algorithm to be executed in parallel on a large number of computing nodes. The proposed algorithm is verified on a Hadoop cluster. The experimental results demonstrate that the effectiveness of the propose algorithm.