Abstract: Aiming at the green two-echelon vehicle routing problem with fuzzy demand, a hybrid hyper-heuristic algorithm is proposed to solve the problem with the objective of minimizing the sum of vehicle operation cost and fuel consumption cost. Firstly, considering that the solution space of the two-echelon problem is huge and coupled, a clustering decomposition strategy is designed to decompose the problem into several sub problems, so as to reduce the search space of the problem reasonably. Then, an enhanced hyper-heuristic estimation of distribution algorithm (EHHEDA) is proposed to solve each sub problem and obtain the solution of the original problem. Based on the hyper heuristic algorithm framework EHHEDA designs a distribution estimation algorithm based on three-dimensional probability model in the high-level strategy domain, dynamically determines the arrangement (i.e. the high-level individual) composed of search operators in the low-level operation domain, and can effectively control and guide the search behavior of the entire algorithm. At the same time, 10 effective neighborhood search operators are designed in the bottom operation domain, and the simulated annealing mechanism of reheating operation is added as the acceptance criterion of the problem solution (i.e. the bottom individual), which is conducive to the in-depth search in the problem solution space. The simulation experimental results show that the proposed algorithm outperforms the algorithms used to solve similar problems in recent years on most test sets, which verifies the effectiveness of the proposed algorithm.