A planning method for park-level integrated energy system based on system integration theory

System integration has attracted a wide range of interest in various fields to explore synergistic benefits among different independent systems or elements. As a prosperous application, the park-level integrated energy system (PIES) integrates electricity, heat and natural gas supply systems to improve the overall energy utilization efficiency on the user side. However, the planning of the PIES faces challenges such as the interaction of long-term and short-term uncertainties, lack of adaptivity, and improper optimization methods. Therefore, a planning method for the PIES based on system integration theory is proposed in this paper. Firstly, the concepts and elements of system integration theory are defined and introduced. On this basis, an integration framework for the PIES planning is constructed, which coordinates dynamic planning, evaluation and local adjustment during the whole integration time. Then, integration optimization models consisting of the global optimization model (OPT3), stage optimization model (OPT2), and local optimization model (OPT1) are established, which work collaboratively under the integration framework. OPT3 governs the entire integration process to achieve the maximum integration value. At the beginning of each stage, OPT2 optimizes the investment and operation strategies. Furthermore, at each local adjustment point within the stage, OPT1 evaluates the current integration scheme and optimizes adjustment strategies. OPT2 and OPT1 are called repeatedly in a rolling way to update the integration scheme dynamically with the progress of integration, reducing the impact of long-term uncertainties. At the same time, to handle short-term uncertainties of multi-energy loads and renewable energy sources, appropriate optimization methods are selected adaptively to solve OPT2 and OPT1 according to actual conditions and requirements. The deterministic optimization, stochastic optimization and robust optimization-based OPT2 and OPT1 are constructed for demonstration. Finally, the effectiveness of the proposed method is verified. The simulation results demonstrate that the proposed method obtains a more cost-effective scheme compared to the multi-stage-based and multi-stage rolling-based methods, with reductions of 5.23 % and 4.25 %, respectively.