Research on the day-ahead scheduling optimization method of medium-depth geothermal cascade heating system

Abstract

The medium-depth geothermal cascade utilization technology plays a crucial role in achieving low-carbon and low-cost heat supply. However, the traditional rigid regulation method faces challenges in ensuring real-time supply-demand balance due to the district heating demand being influenced by outdoor weather and the complexity of the operating parameters of the medium-depth geothermal cascade heating system. Consequently, problems such as low utilization of geothermal water and high operating costs arise. To address these issues, this study proposes a day-ahead scheduling optimization method for the medium-depth geothermal cascade heating system, integrating the adaptive moment estimation (Adam) optimized long short-term memory (LSTM) prediction algorithm, interior point method, and particle swarm optimization (PSO) algorithm. Based on enhanced precision prediction, this study aims to optimize variables including partial load rate (PLR), flow rate, cooling water return temperature, and chilled water return temperature. The effectiveness of the proposed optimization method is verified through the establishment of a TRNSYS simulation platform in conjunction with MATLAB, focusing on a geothermal cascade heating project in Tianjin. The results demonstrate that the proposed day-ahead scheduling optimization method effectively promotes the coordinated operation among energy systems, achieving significant energy savings for the energy station (10.13 %). The novelty of this research lies in the utilization of scientific algorithms to optimize multiple variables within the system, which holds the potential to enhance the matching of supply and demand and improve the operational effectiveness of comparable systems. In summary, this study provides valuable insights and practical implications for the field of strategic optimization. However, it is essential to acknowledge that the applicability of the findings to other geographic locations, different scale projects, or varying system configurations necessitates additional investigation. The generalizability is subject to further research and exploration.