| With the continuous development of robot technology,the rigid linkage and ontology structure of traditional rigid robots greatly limit the ability of direct dynamic interaction between traditional robots and the environment.In recent years,soft robot has become one of the hot spots in the field of robot research.The soft crawling robot inspired by biology has attracted more and more attention because of its ability to complete various complex environmental tasks.In the interaction process with complex environment,compared with rigid structure robot,soft crawling robot has the advantages of high degree of freedom,strong adaptability to complex environment,and high human-computer interaction security.Therefore,it is necessary to conduct in-depth research on intelligent soft crawling robot oriented to environment interaction.This paper designs and manufactures an intelligent soft crawling robot system for environmental interaction.The main research contents include:1)design and manufacture of intelligent soft crawling robot;2)mathematical modeling and fluid-structure coupling simulation of soft crawling robot;3)design and research of the interaction module between the soft crawling robot and the environments;4)build a software crawling robot drive control platform for environmental interaction experiments.In this paper,the body of soft crawling robot is designed to simulate the motion of earthworm.The combination of ontology,flexible bending sensor,infrared digital obstacle avoidance sensor and ORB-SLAM algorithm framework enables the soft crawling robot to intelligently pass through maze channels,complete the environment mapping function in a closed environment,obtain accurate map information,and complete the interaction with the environment.This article mainly carried out the following research contents.(1)Design and manufacture of soft crawling robotAiming at the movement demand of soft crawling robot and the problem of low bearing capacity of flexible materials,a fiber reinforced double cavity flexible actuator and a positive pressure anchor adsorption actuator were designed,and the actuator was manufactured by high-precision 3D printer mold casting.The double cavity flexible actuator is embedded with Kevlar fibers to limit its radial deformation and enhance its stiffness.By switching the inflatable state of single and double cavities,the actuator can accurately complete the bending and stretching action to imitate the movement form of animals.The adsorption actuator is inspired by the octopus sucker,and consists of a double-layer structure with an embedded spiral pneumatic channel at the top.After positive pressure drive,the top expands into a 3D dome structure,which drives the deformation of the substrate film to form negative pressure in the bottom cavity,and the adsorption is anchored on the plane.The double cavity flexible actuator is combined with the adsorption actuator to form a complete body of soft crawling robot.(2)Modeling and simulation of soft crawling robotIn order to verify the rationality of the design and show that the soft crawling robot body can meet its motion requirements when interacting with the environment,a segmented constant curvature model is used to establish a single-cavity pressure path for the dual-cavity flexible actuator of the soft crawling robot torso.To establish a mathematical model of fiber constraint,establish a double-cavity isobaric mathematical model based on the geometric relationship of the actuator design parameters;use a simplified approximate double-layer model to establish a double-layer 3D dome structure mathematics for the adhesion actuators of the head and tail of the soft crawling robot model.(3)Environmental interaction system of soft crawling robotIn response to the interaction between the soft crawling robot and the environment,a flexible bending sensor based on the sensitive grid structure is designed to monitor the bending angle of the robot’s torso in real time.It is innovatively manufactured using the magnetic printing method to complete the performance test and verify the sensor suitable for real-time detection feedback software crawling robot trunk bending Angle,the resistance signal of sensor is converted into digital signal which can be recognized by the control board through the sensing circuit.W503-13 infrared digital obstacle avoidance sensor and flexible bending sensor were selected to realize the autonomous obstacle avoidance function of the robot.A hardware platform was built to calibrate the data of the obstacle avoidance sensor and fix the detection distance to detect the influence of the target’s reflectivity on the detection effect.The framework of feature point method ORB-SLAM is studied using Zhang Zhengyou checkerboard calibration method.The calibration of wireless monocular camera provides support for the subsequent environment mapping by SLAM.(4)Experimental study on environmental Interaction of soft crawling robotAiming at the driving requirements of the soft robot,a pneumatic control platform based on Arduino development board was designed for the crawling of the soft robot.The flexible bending sensor and infrared digital obstacle avoidance sensor were combined to enable the soft robot to complete the crawling motion intelligently and realize the obstacle avoidance function in a closed environment.After crawling,the basic structure of the environment map is obtained by using the video shot by the wireless monocular camera and the framework of THE ORB-SLAM algorithm.Based on Labview and NI data acquisition card,the data acquisition and display program is compiled,combined with the sensor information collected in real time by Labview data acquisition platform and the crawling speed of the robot.The accurate size of the environment map is obtained to complete the reconstruction of the closed map.The size error between the built map and the actual environment is less than 10mm,which verifies the soft crawling robot’s ability to interact with the closed environment. |
PDF Full Text Request