A baseline-oriented dynamic aggregation approach for demand-side heterogeneous controllable resources.
Abstract
The vast number and wide geographical distribution of controllable resources on the demand side make their direct participation in demand response (DR) programs challenging. To improve the coordination of demand-side resources in power systems, this paper proposes a dynamic aggregation method for heterogeneous controllable resources based on load baselines. The proposed framework includes quantitative evaluation of resource response potential, generation of target load baselines, and dynamic aggregation guided by indicator weighting. Response capability indicators are used as clustering features, and the load baseline is segmented into ramping, peak, and valley periods through edge-point detection. To accommodate different regulation requirements, a combined Analytic Hierarchy Process and Anti-entropy Weight Method (AHP-AWM) is employed to determine dynamic indicator weights. An improved K-means algorithm incorporating these weights is used to perform adaptive aggregation of controllable resources under multiple regulation scenarios. Numerical simulations demonstrate the applicability of the proposed framework and show that the resulting clusters exhibit distinct response characteristics under different operating conditions.