模型不确定性下两层电热水器系统的高效实时成本优化,Energy Conversion and Management

---

Efficient real-time cost optimization of a two-layer electric water heater system under model uncertainties

The objective of this paper is to implement an automated system for the control of all energy resources and devices within smart buildings. The paper develops a Two-Layer Electric Water Heater Management (TL-EWHM) system tailored for efficient power distribution within a smart microgrid connected to both a photovoltaic (PV) system and the Electric Grid (EG). The TL-EWHM consists of two distinct layers: the Offline Optimization Layer (OFF-OL) and the Online Optimization Layer (ON-OL). Within the OFF-OL, an optimization algorithm known as Particle Swarm Optimization (PSO) is employed to determine the optimal power settings for Electric Water Heaters (EWH) based on forecasts of several factors including PV generation, ambient temperature (), water demand (), and the consumption of fixed appliances () for the upcoming 24 h. These parameters are forecasted using a novel prediction algorithm referred to as regressor stacking, which is developed based on actual data. Within the ON-OL, real-time data is effectively utilized, and an Extremum-Seeking (ES) Controller is employed to continuously adapt the EWH power settings in response to uncertainties in the predictions. Furthermore, a strategic approach is being devised to generate the water temperature setpoints () by considering dynamic data factors like power, total energy consumption (), and the water demand profile to ensure that user comfort is maintained. A comprehensive comparative analysis is carried out to evaluate the effectiveness of our strategy approach. This strategy is characterized by the inclusion of a Water Temperature Setpoint Generator () and the integration of a two-layer optimization methodology that combines the OFF-OL and the ON-OL. In this comparative study, our strategy is assessed in comparison to other conventional approaches, with the objective of highlighting the distinctive advantages of our method in terms of cost reduction and the enhancement of user comfort. On the one hand, comparisons are made with strategies that give priority to offline optimization. On the other hand, the strategy that maintains a constant water temperature setpoint throughout the day. The results of this paper establish the effectiveness of the proposed strategy. In particular, the energy cost savings associated with the two-layer optimization approach is 13.73%. Even more noteworthy is the significant reduction in energy costs, which reaches an impressive 18.63% when the variation of water temperature setpoints is taken into account. These results underscore the significant benefits of using a two-stage optimization strategy over relying solely on offline optimization. They also underscore the value of incorporating a WTSG as opposed to maintaining WTSC.