Abstract The optimal design of microgrids with thermal energy system requires optimization techniques that can provide investment and scheduling of the technology portfolio involved. In the modeling of such systems with seasonal storage capability, the two main challenges include the low temporal resolution of available data and the non-linear cost versus capacity relationship of solar thermal and heat storage technologies. This work overcomes these challenges by developing two different optimization models based on mixed-integer linear programming with objectives to minimize the total energy costs and carbon dioxide emissions. Piecewise affine functions are used to approximate the non-linear cost versus capacity behavior. The developed methods are applied to the optimal planning of a case study in Austria. The results of the models are compared based on the accuracy and real-time performance together with the impact of piecewise affine cost functions versus non-piecewise affine fixed cost functions. The results show that the investment decisions of both models are in good agreement with each other while the computational time for the 8760-h based model is significantly greater than the model having three representative periods. The models with piecewise affine cost functions show larger capacities of technologies than non-piecewise affine fixed cost function based models.

中文翻译：

---

具有季节性存储和分段仿射成本函数的微电网中热能系统的优化规划

摘要 具有热能系统的微电网的优化设计需要优化技术，可以提供所涉及的技术组合的投资和调度。在对具有季节性存储能力的此类系统进行建模时，两个主要挑战包括可用数据的低时间分辨率以及太阳能热和热存储技术的非线性成本与容量关系。这项工作通过开发基于混合整数线性规划的两种不同优化模型来克服这些挑战，其目标是最小化总能源成本和二氧化碳排放量。分段仿射函数用于近似非线性成本与容量行为。开发的方法应用于奥地利案例研究的最佳规划。基于精度和实时性能以及分段仿射成本函数与非分段仿射固定成本函数的影响来比较模型的结果。结果表明，两种模型的投资决策相互吻合，而基于 8760-h 模型的计算时间明显大于具有三个代表性时期的模型。具有分段仿射成本函数的模型比基于非分段仿射固定成本函数的模型显示出更大的技术能力。结果表明，两种模型的投资决策相互吻合，而基于 8760-h 模型的计算时间明显大于具有三个代表性时期的模型。具有分段仿射成本函数的模型比基于非分段仿射固定成本函数的模型显示出更大的技术能力。结果表明，两种模型的投资决策相互吻合，而基于 8760-h 模型的计算时间明显大于具有三个代表性时期的模型。具有分段仿射成本函数的模型比基于非分段仿射固定成本函数的模型显示出更大的技术能力。