A multi-objective integrated optimal online trajectory planning method for citrus picking robots

Abstract In order to improve the operation efficiency and impact resistance of the robotic arm in unstructured environments such as orchards, this paper proposes an online trajectory planning method for multi-index comprehensive optimization. First, the kinematics and path planning model of the robotic arm is established; then, the joint motion trajectory is constructed using seven polynomials, and a multi-objective comprehensive trajectory optimization model is established using the nonlinear planning method based on kinematic constraints. Secondly, after obtaining the mathematical relationship of each optimization target through analysis, reconstruct the model, simplify the kinematic constraints by using the extreme value principle, and then define the target function by the weighting coefficient method to achieve the comprehensive optimization of the working time, energy consumption and acceleration of the robotic arm motion trajectory. Finally, use the algorithm to optimize the target function to obtain the comprehensive optimal motion trajectory, and take the citrus picking operation as an example for simulation and experimental testing. The simulation results show that the average calculation time of the algorithm is 61 ms, the time index of the obtained comprehensive optimal trajectory is 15.78 s, the energy consumption index is 188.86 °/s², the acceleration index is 405.08 °/s³, and the speed, acceleration and acceleration of each joint are all satisfied. Kinematic constraints. The experimental results show that the citrus picking robot arm runs smoothly and can meet the picking requirements. The average picking time for a single fruit is 16.4 s. This research provides a new method for the online trajectory planning of picking operations with high computational efficiency and can achieve multi-objective comprehensive optimization.