Abstract: Aiming at the green path planning problem of picking and delivering automated guided vehicles (AGVs) with time windows constraints in the context of flexible manufacturing workshops, minimizing the energy consumption and time deviation energy consumption of AGV collection and distribution process as the combined optimization goal, the AGV green vehicle path planning model is constructed. According to the characteristics of the research problem, a hybrid genetic algorithm with improved variable neighborhood search (GA-VNS) is proposed to solve it, a series of five neighborhood structures are designed to improve the algorithm's optimization ability. The feasibility of the proposed algorithm in this paper is verified by solving the test set of Solomon benchmark and comparing the solution with the internationally best-known optimal solutions. Further, take the AGV logistics transportation task in a certain production period of a flexible manufacturing workshop as an experimental case, the algorithm designed in this paper, GA, and VNS algorithm are adopted respectively. Through a detailed analysis of experimental results, the optimization and applicability of the model and algorithm proposed in this paper are verified. It provides a feasible solution for the workshop to achieve the development goal of energy saving and emission reduction.