Aplicación de algoritmos de agrupamiento para la optimización de tiempos de ejecución y eficiencia en algoritmos relacionados a problemas de planificación de sistemas de suministro de energía eléctrica

Abstract

In the context of increasing electrical demand and the integration of new technologies, planners must adapt and optimize the grid to ensure efficient and reliable distribution. However, the complexity of so- lution methods used in electric power system planning (EPSP) at all stages can result in excessive execution times. To mitigate these challenges, this study presents a comparison of different clustering algorithms through an approach that restricts the number of groups based on a group weight limit. The evaluated algorithms include hierarchical, partitional, and density-based approaches. The methodology is applied to distribution expansion planning (DEP) at medium voltage and sub-transmission levels th- rough substation expansion planning (SEP), which employs a binary integer linear programming (BILP) model. Additionally, a partitional algorithm is used to address the problem of assigning load nodes to electrical infrastructures. For both contexts, synthetic case studies extracted from specialized literature are adopted. The results demonstrate that these size-restricted clustering techniques significantly opti- mize solution times by reducing the size of the input data set, usually composed of load nodes, through the application of technical criteria.