Research On Control System Design And Path Planning Of Large Warehouse Fire Robot

With the development of mobile robot technology and the firefighting industry,firefighting robots relying on manual remote control are no longer suitable,especially in large warehouses with complex road conditions and a lot of goods.Based on the characteristics of large warehouses,this thesis proposes a control system for a firefighting robot capable of autonomous navigation and fire extinguishing,and conducts improvement research on two commonly used mobile robot navigation algorithms,A* and RRT-Connect.I carried experimental verification out on the firefighting robot.The specific research contents are as follows:1.Based on an investigation of the development status of firefighting robots both domestically and internationally,this thesis analyzes the advantages and disadvantages of commonly used mobile robot chassis structures.Ultimately,the four-wheel differential model is selected as the mobile chassis structure for the firefighting robot.Using Solidworks software,the 3D modeling of the firefighting robot chassis structure is completed,and kinematic analysis of the model structure is conducted.This establishes the relationship between the chassis motion mode and the left and right driving wheels,providing a foundational model for subsequent research on robot path planning methods.2.According to the modular principle,a three-layer structured control system is designed for the firefighting robot.This control system consists of a task management layer,a decisionmaking planning layer,and a motion control layer.The motion control layer,which is based on the STM32F103VET6 chip,is the primary focus of the design.With the support of the attitude sensor and ultrasonic sensor data,this control layer coordinates and controls two BLDC(brushless DC)motors to accomplish the bottom-level motion driving of the robot.The decisionmaking planning layer relies on the Nvidia Jetson Nano B01 mini embedded computer and ROS(Robot Operating System)to run and execute the robot path planning algorithm.It is supported by the laser radar and environment map,enabling the robot to make informed decisions and plan its path accordingly.3.By combining the environmental model of large warehouses and the motion model of firefighting robots,the A* and RRT-Connect global path planning algorithms were analyzed and improved.Simulations were conducted on the Matlab platform to compare the performance of the original and improved algorithms.The simulation results showed that the improved algorithms outperformed their respective original algorithms in all performance indicators,demonstrating the effectiveness of the improved algorithms.In addition,the principle and workflow of the local path planning algorithm DWA were analyzed,and simulations were conducted on the Matlab platform in both static and dynamic obstacle environments.The simulation results showed that the DWA algorithm has good obstacle avoidance ability in path planning in different environments.4.The principles and workflow of the Gmapping mapping algorithm were studied and analyzed,and a navigation function framework for firefighting robots was built in the ROS environment.Based on the function package,the Navigation navigation function was mainly divided into five parts: map server,costmap＿2d,localization,nav＿core,and move＿base.The firefighting robot’s environment was modeled based on the laser radar,and simulations were conducted in the ROS environment to verify the proposed improved path planning algorithm for firefighting robots.The visualization of the robot’s simplified diagram and running process was achieved through the rviz platform.5.A fire-fighting robot experimental prototype was developed,and a wide corridor in the laboratory and a narrow and complex environment in the laboratory were used to simulate the actual scenario of a large warehouse.The fire-fighting robot was used in two different environments to complete mapping and navigation experiments.The experimental results showed that the fire-fighting robot could plan a reasonable path in different environments and achieve autonomous navigation,which verified the effectiveness of the control system and algorithm improvements designed in this thesis.