Research on multi-timescale coordinated operation methods for long-term physico-electrochemical hybrid energy storage systems

The large-scale integration of a high proportion of new energy sources into the grid presents significant challenges for power supply guarantee and renewable energy consumption. In this context, this paper constructs a hybrid energy storage system (HESS) consisting of long-duration physical energy storage and electrochemical energy storage, specifically compressed air energy storage (CAES), lithium iron phosphate batteries, and supercapacitors, and proposes a multi-timescale coordinated operation strategy for this system. Firstly, the unified energy-circuit modeling theory and the heat loss situation in actual operation of CAES are studied in depth, leading to the construction of an equivalent circuit model for CAES. This makes the overall HESS model more suitable for energy management scenarios. Subsequently, to match various energy storage systems with power frequency bands of different time scales, a hierarchical multi-time-scale coordinated optimization operation method is proposed. The upper layer, through rolling optimization combining day-ahead prediction and intra-day correction, formulates energy management plans on a longer time scale. The lower layer, through power distribution and rule-based strategies, achieves real-time control to maintain instantaneous stability on a shorter time scale. Multi-time-scale coordinated control of the HESS is realized through multi-objective optimization, small-time-scale correction, and deviation feedback. The scientific innovation of this study lies in: constructing a unified modeling approach suitable for the coordinated control of hybrid energy storage systems; and proposing a coordinated control strategy for HESS that incorporates a long-short timescale closed-loop mechanism and multi-objective optimization. Finally, measured data from a CAES demonstration project and a region in Northwest China validate the accuracy of the proposed model. The equivalent circuit model of CAES, which considers thermal losses, achieves an State of Charge (SOC) calculation error of only 3%. Furthermore, the proposed multi-timescale coordination strategy reduces operating costs by 35%-47% compared to conventional methods, while better ensuring power supply reliability and renewable energy accommodation.