Two stages K‐means and PSO‐based method for optimal allocation of multiple parallel DRPs application & deployment. Issue 2 (25th February 2020)

• Record Type: Journal Article
• Authors: Ibrahim, Charles; Mougharbel, Imad; Kanaan, Hadi Y.; Abou Daher, Nivine; Georges, Semaan; Saad, Maarouf
• Is Part Of: IET smart grid. Volume 3:Issue 2(2020)
• Page Start: 216
• Page End: 225
• Publication Date: 2020-02-25
• Subjects: particle swarm optimisation -- profitability -- demand side management -- tariffs; PSO‐based method -- optimal allocation -- Demand Response Programmes -- network levels -- incentive‐based DRPs -- customer selection -- multiple parallel DRPs application & deployment; B0260 Optimisation techniques -- B8110B Power system management, operation and economics; Smart power grids -- Periodicals; Computer science -- Periodicals; Energy industries -- Periodicals; Broadcasting -- Periodicals; 333.79110285
• DOI: 10.1049/iet-stg.2019.0019
• Languages: English
• ISSNs: 2515-2947

Abstract

Abstract : Different types of Demand Response Programmes (DRPs) exist and can be simultaneously offered by the electrical utilities through established contracts with customers. Operating simultaneously multiple types of DRPs might lead to undesired results. DRPs might have different responses to objectives, time‐based ones tend to maximise consumption during lowest tariffs periods while incentive‐based ones tend to reduce the usage based on peak events, accordingly contradiction might occur. Thus, synchronising these DRPs and their parameters through an optimised process including customer selection for the appropriate one is a mandatory step. A fair allocation of the various types of DRPs including their execution's priority at a specific time is the main objective of this study. An original approach based on clustering technique for predicting customers' behaviour coupled with a particle swarm optimisation (PSO) to reach an optimal solution for relocation is presented. In this study, an optimal solution is developed; it provides the various DRPs with the most convenient parameters for the best demand/generation balance, utility profit maximisation and operational cost minimisation. The method is validated through a simulation applying a time‐based with two incentive‐based DRPs in the presence of conventional and renewable generation while using Kmeans clustering and PSO on Matlab.