In order to improve the smoothness of the manipulator operation and achieve the precision of the manipulator control, the BP neural network and the ant colony algorithm are combined to realize the manipulator trajectory control. Firstly, the manipulator trajectory model is established, and then the main parameters of the manipulator are trained by the neural network algorithm. Then, the predicted trajectory of the output is compared with the expected trajectory of the manipulator to solve the optimal parameters closer to the expectation. Finally, the neural network model parameters are optimized by the ant colony algorithm. Experiments show that compared with the traditional BP neural network algorithm, the angular displacement of the proposed algorithm is more fitting with the expected angular displacement, and the displacement error in the spatial three-dimensional coordinate system is smaller.