Title: An adaptive power system management with DG placement and cluster-based load forecasting by CS, K-means and ANN algorithms

Authors: A.G. Veeresha; H.A. Maruthi Prasanna; M.V. Likith Kumar; T. Ananthapadmanabha

Addresses: Department Electrical and Electronics, The National Institute of Engineering, Mysore, India ' Karnataka Power Transmission Corporation Limited, Karnataka, India ' Ramiah Institute of Technology, Bengaluru, India ' The National Institute of Engineering, Mysore, India

Abstract: Power system management plays a big task of grid utilisation in the researcher's world with the maintenance of power imbalance by proper power distribution with load forecasting in the network. For better grid utilisation, a distributed generation (DG) and load forecasting system must be integrated to get the reduced power loss. Concern to that placing of DGs is very important for maintaining the good voltage profile. Hence, the paper proposed a novel system for complete power system management with proper grid utilisation based on DG placement and load forecasting using cuckoo search (CS) and K-mean with artificial neural network (ANN). The proposed system placed the DG in a most appropriate place by optimisation subsequently predict the load by utilising cluster-based algorithm thereby reducing the power loss and maintain good voltage profile. The proposed system is tested in standard IEEE-30 distributed bus system suggested that the performances analysed based on power qualities as well as load forecasting provides better performance on power qualities with minimum power loss of 0.035 kW and the good total voltage profile of 312 kv.

Keywords: distribution generation; DG; placement; cuckoo search algorithm; artificial neural network; ANN; load forecasting; power grid utilisation.

DOI: 10.1504/IJPELEC.2021.114466

International Journal of Power Electronics, 2021 Vol.13 No.3, pp.380 - 398

Received: 19 Mar 2018
Accepted: 12 Nov 2018
Published online: 23 Apr 2021 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article