A Partition Optimization Design Method for a Regional Integrated Energy System Based on a Clustering Algorithm

Abstract This study proposes a partition optimization design method that combines K-means and genetic algorithm (GA) for regional integrated energy systems (IESs) composed of complex loads. Building partitions are obtained by a multi-energy unified clustering model based on a K-means algorithm and performed for the electrical, heating, cooling, and gas loads of various buildings. The availability of all types of resources is evaluated by clustering the local resource data. According to the unified analysis of characteristics such as the thermoelectric ratio of loads and availability of resources, suitable energy supply equipment are selected to form an alternative structure set. Capacity configuration optimization models are established based on energy, economy, and environmental evaluation indicators considering the off-design performance of the equipment. The GA is used to optimize the configuration of each alternative structure for every partition. The system evaluation indicators are sorted by the linear weighting method to obtain an optimal system configuration suitable for each building partition. A case study is used to verify the effectiveness of the proposed method. The energy supply effect of the system designed for each partition is significantly improved compared with that of a separated production system. The proposed method has engineering significance for guiding the construction of regional IESs.