Research On Multi-state Artificial Muscles Control Method Based On Stochastic Recruitment

With the rapid development of research and development in the field of robot control,artificial muscle compliance control has gradually become a hot spot in the field of robot research.Compared with traditional electromagnetic actuators,robotic actuators made of artificial muscle materials have the advantages of high degree of freedom,high energy density,low noise,and can be close to the effects that biological muscles can achieve.However,materials constituting artificial muscle actuators such as shape memory alloys(SMA)have inherent intrinsic characteristics,severe hysteresis and nonlinearity,and control is more difficult than conventional electromagnetic actuators.Therefore,it has theoretical and practical significance for artificial muscle control methods and model research.Based on the two-state artificial muscle cell,this paper mainly analyzes and studies some control algorithms and actuator architecture problems involved in the polymorphic artificial muscle control system based on random enlistment.The main work and research results are as follows:Firstly,by analyzing the characteristics of SMA materials and using the idea of stochastic recruitment control,a closed-loop control law for stochastic recruitment of multi-state failure cells is established for the problem of multi-state failure cells,and the performance of the system is analyzed by Simulink simulation platform.In the case of the overshoot problem of the stochastic recruitment control system,the failure cell selection controller is added,and the stochastic recruitment control system without overshoot failure cell is established,which solves the overshoot problem in the system.Secondly,from the perspective of bionic control,the physiological basis of limited feedback control is explained,and the multi-state cell limited feedback stochastic recruitment control system is established.For the two-state cellular and multi-state cellular systems,the derivation of the limited feedback control law is carried out,and the two kinds of limited feedback control algorithms are simulated and compared in the Simulink environment.Aiming at the key scaling factor problem in the system,the longitudinal and lateral contrast experiments were carried out to analyze the influence of the scaling factor on the system.Thirdly,the problem of the adjustment time of the multi-state cell stochastic recruitment control system is studied,and the feasibility of applying the optimal control theory to the artificial muscle control system is explored.For the two-state cell and multi-state cell system,the optimal control law of artificial muscle based on dynamic programming method is established respectively.The simulation and comparison analysis of the two control algorithms are carried out in Simulink environment.Finally,an antagonistic artificial muscle microactuator system based on stochastic recruitment was established.From the two aspects of bionic control and artificial muscle selection,the artificial muscle antagonist microactuator system model was demonstrated and constructed,and the actual parameters were used to derive the range of angles that the actuator can perform.The traditional closed-loop controller and optimal controller are applied to the artificial muscle antagonist microactuator system.The simulation design shows the rationality and effectiveness of the controller design.