Inverse Kinematics Solving of a Novel Soft Manipulator with Vision and IMU Data

Currently, robotic manipulators are more and more important in the application of service robots. Compared with traditional rigid manipulators, the flexibility of soft robotic manipulators reduces the risk of human injury and damage to delicate objects during manipulation and make them more adaptive to some complex environment. Therefore, the soft manipulator shows better application potential in rescue, disaster relief, and human-robot interaction tasks. However, soft manipulators often have redundant degrees of freedom. So the accurate model and control of the soft manipulator based on rigid body dynamics are still challenging. In this paper, we design a novel soft manipulator, which is driven by nine steering gears with stainless steel wire ropes and has high control precision. Firstly, we analyzed its kinematics of the manipulator and then built a multi-data acquisition system based on the vision positioning system and inertial measurement unit. When the soft manipulator performs the periodic movement, this system can record the coordinate of the end position, as well as the attitude information and time stamp in real time. Subsequently, the inverse kinematics solving method can be transformed into the optimization process using the acquired data. Experimental results show that we can achieve the control accuracy of 5mm to 7mm, which proves the effectiveness of our data-driven inverse kinematic solving method.