Comprehensive evaluation of low voltage station area is essential for the operation and monitoring of smart distribution network. The existing comprehensive evaluation methods of the low voltage station area is very subjective, and the factors affecting the comprehensive evaluation are too many to lead to the low accuracy of the evaluation model. In order to solve this problem, Comprehensive evaluation algorithm of low voltage station area in smart distribution network based on hybrid gene expression programming is proposed which combines with rough set theory. Firstly, based on the idea of rough set, the optimal reduction is solved by the calculation of the dependence degree of the comprehensive evaluation decision table. Secondly, comprehensive evaluation function of low voltage station area in distribution network is mined by gene expression programming. Finally, based on the function model, the purpose of the evaluation and prediction of the low voltage station area operation state is evaluated. Experimental results show that comprehensive evaluation function model on distribution low voltage station area based on the proposed algorithm has high efficiency of function mining, accuracy of prediction and strong practicality.