Modeling And Control Of An Energy-efficient Actuator For A Biomimetic Flexible Robotic Joint

Robots play a more and more important role in national economy,medical health,national defense and so on.Although the research on robots has made great progress,the problem of unnatural movements and great energy consumption,because of the rigid structure in most robotic joints,restricts the development and application of robots.The human body can improve the efficiency of energy utilization in the process of movements for the flexible units in the musculoskeletal system.In this paper,an energy-efficient actuator for biomimetic flexible robotic joint based on biomechanical characteristics,strategy for parameter optimization according to the energy criterion and the controller of the biomimetic flexible joint are established to improve the energy efficiency.The main research issues and results are summarized as follows.First,in order to improve the energy utilization rate,the paper proposes a biomimetic flexible actuator.The mathematical model of the flexible actuator is established to analyze the energy consumption of the robotic joint theoretically.Second,aiming at solving the problem that the energy index of flexible actuator is difficult to be quantified,a method of frequency domain synthesis of planned motion curve is proposed.Through the Fourier transform and according to the energy criterion,the parameters of the flexible actuator are optimized by genetic algorithm.An optimization strategy for variable stiffness system is proposed to further improve the energy efficiency and adaptability of the robotic joints.Third,aiming at solving the problem that the control performance and the energy index of the flexible joint are difficult to be coordinated,a feedforward-feedback-virtual damping compensation motion controller of the biomimetic flexible joint is designed,based on trajectory tracking and energy optimization targets.The virtual damping compensation controller and the feed-forward predictive controller based on linear model are added to reduce the energy consumption of error regulations.The PID feedback controller based on BP neural network is designed to enhance the robustness of the system.Finally,a virtual flexible joint model is built in a dynamics simulation environment based on Adams.The presented theories are verified by combined simulation methods.Through the analysis of the energy dissipation and power value,the advantage on energy efficiency of the biomimetic flexible actuator proposed in this paper is verified.