1.School of Mechanical Engineering and Automation, Zhejiang Sci-Tech University, Hangzhou 310018, China;2.College of Electrical Engineering, Zhejiang University, Hangzhou 310058, China

The existing recommendation methods for the electricity retail plans based on the evaluation information of customers ignore the diversity of differentiated customer evaluation information and only consider the cases, in which customers are good at evaluating all the attributes of the electricity sales, which will bring great deviation to the recommendation results. To this end, a recommendation method for the electricity retail plans based on the bi-level affinity propagation (BLAP) clustering algorithm and multigranular hesitant fuzzy linguistic evaluation set is proposed. Firstly, the division method for the sample customer set based on the customer portrait label system and BLAP clustering algorithm is proposed to identify the customers with similar electricity consumption characteristics. Secondly, considering the multigranular hesitant fuzzy linguistic evaluation set and incomplete weight information, a fuzzy evaluation method for the selection of electricity retail plan by the sample customer set is put forward. Thirdly, based on the information of the sample customer set, the new evaluation method of new customers and the full-ranking recommendation method for the electricity retail plan are proposed, which can realize the accurate recommendation of electricity retail plans by electricity retail companies. Finally, taking customers in an area of China as a case for study, the results show that the electricity retail plan recommendation method for electricity retail plans based on BLAP clustering and multigranular hesitant fuzzy set can help electricity retail companies improve the recommendation quality. Thus, the satisfaction of customers is improved, and the stickiness of customers is enhanced.