基于液-气转化机制软体机械手的设计与位置控制

Design and Control of Soft Manipulator Based on the Liquid-vapor Transition Mechanism

针对现有软体机械手存在形变模型不准确、操作空间小和位置控制精度低等问题,设计出一种液-气相变驱动三关节软体机械手并进行了相应的形变改进模型分析与运动控制试验。建立一种考虑温度与材料自重因素的硅胶形变非线性力学模型且基于该模型完成了机械手的弯曲性能分析,建立三关节机械手末端操作空间的数学模型。而后基于传统刚性机械手的运动学逆解算法,提出一种基于最小驱动能量函数的运动学逆解方法来确定机械手各关节的弯曲角度。通过最小驱动能量函数并利用神经网络算法策略实现了三关节软体机械手末端轨迹的稳定控制,其相对位置距离偏差小于2.8%。通过充分的试验和理论分析,为液-气相变转化驱动软体机械手的后期实际应用提供了理论支撑。

Traditional soft manipulators demonstrate several shortcomings, such as small working space, inaccurate deformation model and position tracking control. In this paper, a three-joint soft manipulator based on liquid-vapor actuator is fabricated, modeled and tested. Firstly, a nonlinear mechanical model considering the gravity and temperature of the silicone material is proposed to achieve accurate bending model of the manipulator. The working space of the manipulator is analyzed using D-H coordinate method.Secondly, a kinematics inverse solution method based on the minimum actuating energy function is used to determine the bending angel of each join for the manipulator. Finally, a hybrid control strategy by combining the minimum actuating energy function and radial basis function neural network is developed to control the position of manipulator. The experimental results demonstrate that the proposed method leads to accurate position tracking results. In conclusion, the design and control method proposed in this paper may provide an alternative to soft manipulators in the future.