Optimization Design And State Estimation Of A Pneumatic Square Soft Manipulator

The soft manipulator has the characteristics of self-adaptive flexibility,friendly human-computer interaction,and strong environmental adaptability.It is a powerful way to solve many problems in unstructured environments and complex working conditions,with great potential and development prospects.The soft manipulator has great potential and development prospects.This paper takes a pneumatic square(PnS)soft manipulator as the research object,and carries out research work around the optimization design,molding manufacturing,and state estimation of the manipulator.The research results obtained in this paper are as follows:1.Optimal design of the PnS soft manipulator.Taking the bending stiffness of the manipulator as the evaluation index,the relationship between the parameters and the structural performance of the cavity is obtained based on the finite element simulation;The Latin hypercube sampling method was used to draw samples.A proxy model between the design parameters and bending stiffness of the PnS soft manipulator was constructed.In a given design space,the best combination of design parameters is obtained with the maximum bending stiffness as the goal by the search algorithms2.Manufacturing process optimization of the PnS soft manipulator.Based on the PnS soft manipulator model,the 3D printing manufacturing method has been research to solve the manufacturing problem of the manipulator.Considering the 3D printing manufacturing method and the structural characteristics of the PnS soft manipulator,methods for optimizing the manufacturing model are proposed,such as expanding the bottom surface edge and thickening the top surface to compensate for losses.These methods significantly improve the manufacturing quality and realize the fusion of material optimization,structural design and manufacturing methods.Exploring the manufacturing application of multi-nozzle printing equipment,to realize the rigid-soft integrated manufacturing of the manipulator,and then build a manipulator experimental platform.3.Bending state estimation of the PnS soft manipulator.Perform experiments on the PnS soft manipulator designed and manufactured above,use image processing technology to obtain the coordinate values of the marker points of the manipulator,then unify the extracted coordinate data to construct the initial data set.The bending curve of the manipulator was fit by a second-order polynomial.The curvature value of the center point is used to describe its bending state.Based on the neural network model,the multi-layer perceptron model between the input of the manipulator and the coordinates of the marker points is trained to realize the state estimation of the PnS soft manipulator under the no-load working condition.The research results of this paper are of great significance to the development of structural design,rapid manufacturing,state perception and practical applications of soft manipulators.