Design Of Remote Operation Management And Decision-making Planning System For Inspection Robots

The mobile inspection robot,as a branch of robots,combines the Internet of Things and Artificial Intelligence Technology to improve the efficiency of robot inspection,which is developing in the direction of intelligence.However,the inspection robot still has some problems,such as the limited functions under remote monitoring,excessive reliance on manual maintenance,high costs on operation,et al.Holding the concept of human-machine cooperation,this study focuses on the needs of inspection tasks and designs a remote operation management system for the inspection robot with a hierarchical inspection decision planning system based on the behavior tree.This study combines MQTT communication,streaming media technology,behavior decisionmaking,path planning and other technologies in the ROS environment.The effectiveness and applicability of the proposed system are verified by experiments in real-world scenarios.The main research work and results of this thesis include:1.A lightweight inspected data communication system with superior performance,lower latency,and lower costs: through data interaction and storage,streaming media transmission,and other modules,the communication system enables the efficient,safe,and reliable transmission of interactive data,and builds a modular instant communication system that is easy to maintain and extend.The tests show that the system can run stably for a long time,with the data loss rate less than 0.42%,the 1080 Pvideo delay less than 0.6s,and the frame loss rate less than 1.05%.These performances satisfy the requirements of regular operation management.2.To enrich the functions of robots under remote control,this research designs and implements a multi-functional remote intelligent operation management platform for inspection robots.This interactive system has multiple functions,including user login management,interactive data storage,real-time video monitoring,fault detection and alarm,map positioning and navigation,motion command remote control and multi-task data issuance,and other practical functions.3.To meet the needs of the autonomous and intelligent operation of robots,this research designs and implements an autonomous behavior decision planning system based on behavior trees.This system creates a behavioral decision-making mechanism with dynamic adjustment and data sharing functions.Considering the problem of insufficient flexibility of behavior trees,an improved method for selecting node with priority adjustment was proposed.The experiments show that the system has autonomy and intelligence and the robot can autonomously make behavioral decisions based on task data,and dynamically adjust the inspection behavior and feedback the status in real-time.To sum up,the experiments prove that the system has low operating costs,stable data transmission,flexible and effective inspection decisions,and strong applicability.With the layered and modular design,the system manifests stable performance and strong applicability of the robot inspection.