Objective:In order to improve the efficiency of multi-point moving path planning of food picking robot, a path planning method of food picking robot based on density peak clustering parallel sparrow search algorithm is proposed.Methods:The path planning model of food picking robot was established with the total moving distance, path smoothness between points and moving safety as the evaluation indexes. While ensuring the moving safety of the robot, the path smoothness was improved and the moving distance was reduced as much as possible. The density peak clustering sparrow search algorithm (DSSA) was designed, as the improved density peak clustering algorithm was used to cluster the sparrow population, divided different sub populations and defined the sparrow iterative evolution mode according to the clustering results. Combined with the multi-point path planning model and the four potential moving paths between points, the sparrow coding mode was redefined and a parallel computing architecture was build to improve the accuracy and operation efficiency of DSSA solving the path planning model.Results:The simulation results showed that compared with other food robot path planning methods, the total moving distance was reduced by 7.3%~39.2% and the moving time was reduced by 26.7%~50.1%.Conclusion:The proposed method can significantly improve the path planning efficiency of food sorting robot, which has certain application value for improving the production efficiency of food processing enterprises.