Distribution transformer cluster flexible dispatching method based on discrete monkey algorithm

Abstract Distributed generators are being widely integrated into China’s urban distribution network unevenly, which accounts for different load rates of distribution transformer cluster in spatial dimension and low net load simultaneity rate in temporal dimension. Therefore, it is necessary to carry out flexible dispatching for the distribution transformer cluster in active distribution network to effectively reduce the grid loss. Different from the existing research on dynamic reconfiguration of the network supplied by a single transformer or a single feeder, the grid loss of distribution transformer cluster is optimized in the paper. Firstly, a 0-1 integer optimization model is established. The model is characterized by calculating the total grid loss of distribution transformer cluster based on the connectivity of power supply path and considering the spatio-temporal constraints such as total change times of switches. Secondly, uncertainty of renewable energy output is considered with k-means clustering algorithm effectively. The reduced scenarios are used for flexible dispatching simulation. Then, the discrete monkey algorithm, characterized by the cross iteration of local optimization and global search, is improved for solving 0-1 integer programming problem. The improved method can deal with the temporal coupling constraints in the model. Finally, the effectiveness of the proposed method is identified by case study. Flexible dispatching of distribution transformer cluster can greatly reduce the grid loss of active distribution network, and the optimal dispatching scheme will change along with the change of distributed generator output.