| The two basic functions of the wall-climbing robot are adsorption and movement.It is a multi-functional robot used to replace the manual rust removal operation.Since the wall-climbing rust removal robot mainly works on the surface of the hull,and the surface of the hull is complicated,There are not only planes,but also curved surfaces and even obstacles on the surface of the hull.Therefore,it is imperative to study a multifunctional wall-climbing robot that integrates rust removal,obstacle avoidance,automatic operation and path planning.At present,the research on wall-climbing robots mainly focuses on hardware design and research,but ignores the related research on intelligence.In fact,the wall-climbing robot’s automatic obstacle avoidance ability,the adaptive ability of the wall environment,and the data collection of sensors And signal processing needs further research and development.Based on the existing research in the laboratory,this thesis combines the current research status of wall-climbing robots to study the related technologies of automatic control of wall-climbing robots,analyze the robot motion characteristics,establish the robot kinematics model,and base on the wall-climbing robot.The current research status of robot path planning is analyzed to determine the optimization of robot path planning algorithms and other related content.As one of the core technologies of intelligent system design,reinforcement learning methods are widely used in fields such as artificial intelligence and robotics,and autonomous path planning of robots has become a hot topic in the field of robotics.The wall-climbing rust removal robot has the advantages of environmental protection,easy operation,high efficiency,etc,so it has important research and application value in the field of ship hull rust removal.Usually,derusting the hull surface is a very heavy work,so there is an urgent need to improve the automation of wall-climbing rust-removing robots,but the research on wall positioning,route planning,automatic control,etc.has not yet reached the application requirements,such as climbing The variable load characteristics of the wall rust removal robot in the working process bring difficulties to the robot and increase the difficulty of the motion control of the wall-climbing robot.At the same time,the form and distribution of obstacles on the hull will also vary depending on the type of ship,so the wall-climbing robots need to solve problems such as self-positioning,obstacle positioning and route planning in various environments.In this article,we have conducted a detailed study of key technologies based on the relative properties of the robot and the hull wall itself,such as the path planning of the wall-climbing and rust-removing robot,the control system and the control algorithm.At present,the rapid development of robotics technology,artificial intelligence(AI)has also become a research hotspot in recent years.In order to improve the working efficiency of the wall-climbing and rust removal robot,when performing the rust removal work on the hull surface,the robot can ensure that the robot can fully cover the rust and also find the optimal path to reduce energy consumption.For this purpose,a machine learning-based approach is proposed.Research on curved path planning technology of ship hull rust removal robot.According to the different positions and the distribution of obstacles,the problem of wall-climbing robots is solved,and a Q-Learning-based path planning method is proposed.The hull surface discrete model and the robot kinematics model are established,and the surface discrete model is used to establish Local environmental information,through continuous interaction with the environment to obtain environmental information,and then through the feedback enhanced signal to evaluate the selected actions,using continuous trial and error and selection to solve the problem of inaccurate environmental model modeling or unknown obstacles The path planning problem effectively improves the self-learning ability and self-adaptability of the robot system in the path planning problem,so that the mobile robot has been expanded in the path planning field with limited conditions,which has very practical application value. |
