Integrated Navigation Technology Of Inspection Robot In Narrow Environment

At present,inspection robots have been widely used in oil pipeline stations,chemical plants,substations and natural gas stations and other places that require regular inspections.Most of these inspection robots are applied to open and semi-structured environments,which may be filled with various large oil and gas tanks,pipelines,and columns and other facilities,causing some passages to be very narrow,restricting the access of inspection robots.In this paper,a wheeled mobile robot with a four-wheel skid-steering structure is selected as the carrier to study its LIDAR-based navigation technology in a narrow environment.This article takes the gas station inspection as an example.The design of the corresponded chassis size(0.8m × 0.5m × 0.7m),the selection of the driving components,the construct of the LIDAR and the GNSS system are all determined according to the actual situation,and so is the overall design of the inspection robot completed.To improve the control accuracy,the coordinate systems of the LIDAR and GNSS systems were calibrated.According to the actual environment of the station yard,three narrow passages were simulated:column-column passage,column-area passage and area-area passage.LIDAR is used to scan the current environment,obtain potential passages based on the continuity of the scanned data,and determine the passage to be passed in conjunction with the robot's motion direction Search the nearest points on both sides of the passage and this data will be used as a reference to establish a local coordinate system within the passage and determine the robot's pose in the local coordinate system of the passage.The influence of the target distance on the positioning accuracy was analyzed at first to improve the positioning accuracy of the robot,and then the EKF algorithm was applied.Considering that the patrol robot may use GNSS for navigation in an open environment,a federal filter is designed which combines LIDAR positioning information and GNSS information.Combined with the motion mode of the inspection robot in the station yard,the three motion control modes of the robot in polar coordinate system are analyzed:point-to-point motion control,linear tracking motion control and general curve tracking motion control.SIMULINK simulation verified the effectiveness of three motion control algorithms.In order to improve the motion control accuracy of the inspection robot in a narrow environment further,a heading pre-estimation control algorithm is introduced into the motion control model.Simulation results verify that the heading pre-estimation control algorithm has better adaptability at different speeds than PID algorithm,and the control process is smoother and the phenomenon of overshoot is reduced.Corresponding experiments are designed to verify the effectiveness of the point-to-point motion control and heading pre-estimation control algorithm of the inspection robot.Further,the accuracy of the robot as it passes through the area-area passage is analyzed.The experimental results show that when the speed of the inspection robot is 0.2m/s,the positioning error of the inspection robot in the passage is 0.04m,the maximum deviation of the control heading angle is 5 degrees,and it can smoothly pass through the 0.6m wide passage.