Objective:To solve the problems of unstable grasping operation and low sorting efficiency of sorting robot in food production line.Methods:Based on the architecture of high-speed parallel food sorting robot, a multi-objective motion optimization strategy based on improved particle swarm optimization algorithm was proposed for the dynamic target grasping control method of food sorting robot. By coordinating the grasping sequence and sorting trajectory, the shortest path model was established. Establish the mechanism stability optimization model with the end acceleration, and optimized the target by improving the particle swarm optimization algorithm.Results:Through experimental verification, when the conveying speed was 100 mm/s, the grasping success rate was increased from 96.8% to 100%, and the sorting rate was increased from 1.62 to 1.98 s-1.Conclusion:This control method can effectively improve the operation stability and sorting efficiency of the food sorting robot.