Research On Redundant Robot Wireless Control System And Target Recognition For Disaster Environment

The demand for detection and rescue of disaster environments is becoming increasingly urgent.Developing robots to replace humans or help people to carry out these tasks can greatly increase operational efficiency and win valuable rescue time.The operating conditions in a disaster environment are harsh,and robots need to deal with problems such as insufficient light,poor communication,and uncertainty in handling.Based on this,this article takes a seven-degrees-of-freedom redundant robot as object,carries out research on wireless communication control and target identification methods in disaster environment,develops a redundant robot control system based on ROS(Robot Operating System)and carries out experimental verification.Aiming at the content and environment characteristics of robots in disaster environment,a wireless robotized control system with vision is proposed,which realizes the physical separation between the robot motion controller and the operating arm.The control system includes a motion control subsystem,a wireless communication subsystem,and a target perception subsystem.Among them,the motion control subsystem is based on ROS development and integrates functions such as 3D modeling,motion planning,instruction generation,and animation simulation of redundant robots.It can meet tasks before(simulation verification),during execution(real-time control and monitoring),and The different requirements of post-execution(post-processing and task recurring analysis);the wireless communication subsystem establishes the communication link between the motion controller and the robot's joints;the target-aware subsystem identifies the operating object and provides pose information.In order to reduce the use of cables and provide reliable communication,a Wi Fi-based wireless communication controller was developed to realize the data interaction between the motion control subsystem and the robot's joints.The controller obtains the planned motion data from the ROS communication node,sends it to the corresponding joint,and acquires the sensor data from each joint and uploads it to the motion control subsystem.Wireless communication controller uses STM32ZET6 as the main control chip,in addition to Wi Fi wireless communication,it can also realize data monitoring and emergency interrupt control functions.To meet different needs and consider backups,the controller also provides a CAN wired communication interface.Aiming at the typical disposal task requirements in the disaster environment,a kind of valve target recognition and position and attitude measurement methods were proposed in the dim condition to provide environment awareness information for redundant robots to carry out job tasks.Binocular stereo vision is used to obtain target image information,combined with multi-frame images for enhancement,Gaussian filtering is used for denoising,and the image is further subjected to edge detection and feature extraction to obtain valve circumference characteristics and straight line features,after 3D reconstruction.Get the valve target geometry and space pose.Based on the identified information,the trajectory of the redundant robot can be planned.Finally,the sytem integration and experimental research are carried out.The results show that the designed control system is friendly and convenient for human-machine interaction,rich in functions and good in stability.