计及负载变化的容积驱动人工肌肉长度预测方法

The pump driving mode of artificial muscles represents a valveless driving approach, offering advantages such as straightforward driving equipment and high system efficiency. Among the existing length prediction methods for pump-driven artificial muscles, the impact of load force has not been taken into account, resulting in deviations between theoretical and actual lengths, thus urgently demanding improvement. Firstly, an experiment is carried out to explore the influence of load force on the length of artificial muscles. It is discovered that for artificial muscles of varying lengths, the relative length variations caused by the same load force were consistent, but the absolute variations were related to the original length of the muscle. Specifically, the longer the original length of the artificial muscle, the greater the absolute length variation induced by the load force. Hence, for longer artificial muscles, the influence of load force must be considered. Subsequently, the relationship between the external load force and the length of artificial muscles is obtained through experiments, and based on this, the original pure geometric length prediction model is revised. Finally, it is hypothesized that the state variables of artificial muscles, such as the length of unloaded muscles, the muscle elastic modulus, the pressure-load force correlation coefficient, etc., are all functions of the injected volume. Thus, a length prediction method for pump-driven artificial muscles is established with the injected volume as the core variable, and verified through single artificial muscle experiments and parallel mechanism application experiments. The experiments indicated that this method can achieve considerable accuracy even under open-loop control conditions without position feedback. This provides a new technical approach for the application of pump-driven artificial muscles in the field of robotics.