The work aims to improve the working efficiency of palletizing robot, reduce the energy loss, and optimize the work trajectory of robot end hand. Firstly, the mathematical model of robot path planning was established. On the basis of this, the traditional ant colony algorithm was improved. The local robot path information in the environment was introduced into the initialization of ant pheromone and the path selection probability, to improve the convergence speed of ant colony algorithm and prevent the algorithm from precocity, and protect the algorithm from falling into local optimum. The simulation results showed that, the improved ant colony algorithm converged faster and could plan the optimal path in a short time. The proposed method can obviously improve the optimal path searching ability of the palletizing robot, and it is of great significance to improve the operation efficiency of the robot.