The spatial flexible robotic arm grasps mass perturbations during the operation task, causing changes in the force and deformation related parameters of the dynamic model, making the dynamic model full of uncertainty, and thus causing deviations in the operating position of the robotic arm. The solution of control parameters is prone to getting stuck in local optima, seriously affecting control accuracy. Therefore, a spatial flexible robotic arm visual feedback steady-state control system based on hybrid multi strategy improved beetle algorithm is proposed. In terms of hardware, design visual sensors, motion control cards, and servo motor drivers. In terms of software, considering the kinetic energy, elastic force, and generalized force of flexible bodies, a spatial flexible robotic arm dynamic model is established to reflect the force situation and deformation characteristics of the robotic arm during motion, as the basis for steady-state control. By using a large number of images collected by visual sensors, image visual feedback information and position visual feedback information are obtained. Based on this, the feature error and position error generated by the motion of the robotic arm are calculated, and the motion state tracking of the robotic arm is completed. Build a visual feedback steady-state control framework for a robotic arm, input the motion state tracking results into it to derive the steady-state control law, and apply it to the dynamic model to complete the steady-state control processing. Using a hybrid multi strategy approach that combines chaos mapping strategy and artificial jellyfish search algorithm to improve the beetle algorithm, the steady-state control gain parameters are tuned to achieve better steady-state control results for the system. The test results show that after using the system to complete steady-state control processing, the motion attitude error of the space flexible robotic arm remains within ± 0.2 °, proving the good control performance of the system.