Under the backdrop of the new power system,customer directrix load (CDL)-based demand response,as an efficient demand-side regulation mechanism,has garnered widespread attention.However,existing research primarily focuses on grid dispatch and fails to fully account for the complex challenges faced by virtual power plant (VPP) in decomposing and transmitting macro-CDLs to heterogeneous resources.A CDL decomposition and dispatch method based on users’ electricity consumption characteristics is proposed.Firstly,a resource response feature evaluation system is constructed,where the multidimensional feature extraction and dimensionality reduction are employed to identify electricity consumption patterns,and the clustering algorithms are utilized to classify heterogeneous resources.Subsequently,an optimization model aimed at minimizing regulation costs is established to decompose the macro-CDL into personalized sub-CDLs tailored to different consumption patterns.Simulation results demonstrate that the proposed method effectively reduces regulation costs and improves response accuracy,providing decision-making support for VPP participating in demand response.
