1048

This paper mainly studies the implementation of flower pollination algorithm in multi-objective optimization problems, flower pollination algorithm mostly uses the way of setting weight to transform multi-objective optimization into single objective optimization, but the setting of weight coefficient mostly depends on the experience of experts, it makes the whole algorithm more subjective, the final performance of the algorithm largely depends on the research level of researchers. This paper proposes a multi-objective flower pollination algorithm based on the traditional flower pollination algorithm, which uses non-dominated sorting to find the elite solution of each generation solution set, and then finds the Pareto optimal solution according to the crowding distance. The algorithm is compared with other three kinds of multi-objective optimization algorithms. The results show that multi objective flower pollination algorithm based on non-dominated sorting(NSFPA) has a good effect on solving multi-objective problems.