In order to improve the picking efficiency of outfitting pallets, a mathematical model was established after analyzing the picking process, and an improved genetic algorithm based multi vehicle collaborative picking method for outfitting pallets was proposed. Aiming at the process of standard genetic algorithm, this paper improves its initialization process and chromosome crossover method, and makes modifications to the mutation process that are closer to actual production; Aiming at the problems existing in standard genetic algorithms, this paper adopts a recently proposed variable neighborhood search strategy to reduce the possibility of falling into a local optimal solution; Finally, combining the two parts, an improved genetic algorithm is proposed. The effectiveness of the algorithm is verified by an example calculation, which optimizes the traditional pallet picking method..