A Simultaneous Dual-Cycle Heuristic Algorithm Optimizing Method for Distributed Energy Systems

The distributed energy system emerges as a promising and valuable technology. However, various factors are hindering the development of the algorithm, including the diversity of units and their respective output constraints. Additionally, the multiple-layer algorithm still faces difficulties searching for the best coordination between integer and continuous variables. To address these challenges, this paper proposes a simultaneous optimization method to aid the design of distributed energy systems. Considering the influence of the outputs of the different units, the proposed method introduces a dual-cycle structure that separates the storage energy units and other units according to the output mode. Additionally, the proposed method yields the hourly outputs of the different units and their respective rated capacities simultaneously. At the same time, an originally designed random walk algorithm with compulsive evolution is integrated into the proposed method. Moreover, a time-series optimization method is applied for the storage energy device to enhance the computational efficiency. To validate the proposed method, the configuration of the distributed energy system (DES) and the hourly output of the different units in three scenarios are analyzed in detail. Quantitative results show that the proposed RWCE-DC reduces the average daily total cost compared to a standard differential evolution algorithm with penalty functions (from 413,628 to 241,716 CNY/day in Scenario II). Across three grid-interaction scenarios, RWCE-DC yields daily costs of 243,271 CNY/day (Scenario I), 209,716 CNY/day (Scenario II), and 178,896 CNY/day (Scenario III) while automatically removing redundant units (e.g., gas boiler in Scenario I) and strictly respecting storage state-of-charge constraints without penalty functions. However, the analysis has several limitations. First, the economic model uses a simplified annualized cost approach without taxes, subsidies, inflation, or discount rate variations. Second, only one geographic location with specific solar and load profiles is considered. Third, the current algorithm focuses on single-objective cost minimization and does not yet incorporate multi-objective trade-offs. These factors should be considered when interpreting the absolute cost values and when applying the method to other regions or policy contexts. These results confirm that the proposed dual-cycle method provides an efficient and numerically validated optimization approach for DES synthesis.