The Driving Force Mechanism and Control Strategy of a Pneumatic Hydrostatic Soft Robot

Traditional robots with complicated rigid structure will be worn and teared when running in a cramped and diverse space. For this problem, a new model of pneumatic hydrostatic force soft robot is presented, and the resistance and motion loss are reduced while robots pass the cramped space. Based on sea cucumbers' hydrostatic bone structure, a soft robot with airbag phase transition structure is designed, the hydrostatic mechanical property of the airbag's charging and discharging process is analyzed, and a mathematical model of driving force is set up. Thus, single-unit and double-unit control policies are designed, meanwhile, a time-stamped BP (backpropagation) neural network is used to predict delay time for increasing the stability of internal model feedback control. According to obstacle surmounting tests and stability analysis tests, the robot with delay prediction can move stably and cross small gaps smoothly by single-unit and double-unit control strategy, and the probability of turning over six units increases from 89% to 96%.