Hybrid Ant Colony Optimisation-Random Forest Regression-Based Optimised Framework of An Intelligent Active Distribution Network

Measurement and model uncertainties in power grids lead to the unpredictability of its power output, which brings great challenges to the safe and sustainable system operation. Furthermore, optimal placement strategies of measurement devices and new energy sources have a great impact to maintain the system operation. This study introduces and constructed an active distribution network (ADN) including an energy storage system (ESS) and is optimized by using a hybrid Ant Colony Optimization-Random Forest Regression (HACORFR). Briefly introduces the operation and combined system modelled of the ESS-RDS (ERDS). Then the HACORFR algorithm optimizes the operation status of the RDS and detects the abnormality distribution network. After applying the proposed approach, if the operation indicators of the distribution network, don't return to normal operation, the network topology will further reconstruct and optimize the ERDS. The IEEE-28 bus power distribution testing system was used to validate the proposed method. Finally, the results show that the proposed optimization method reduces the loss of the distribution network by 2.34%, and the voltage profile improved by 2.94%. The present method makes full use of the flexible load and realizes the interaction between supply and demand. It also greatly improves the reliability of the distribution network operation and economy.