Research And Implementation Of Vision/IMU Integrated Attitude Measurement Method For Telescopic Arm Of Boarding Bridges

With the rapid growth of installed wind power,the demand for offshore wind power operation and maintenance is also growing rapidly.As offshore wind turbines are in a more complex and changing environment than onshore wind turbines,the operation and maintenance of offshore wind turbines face more problems,and the control of the boarding bridge telescopic arm and the stable docking of offshore wind turbine landing port is one of the urgent needs of the current operation and maintenance of offshore wind turbines.The purpose of this paper is to study the attitude measurement method of the boarding bridge telescopic arm by combining vision and inertial sensors,and to develop an embedded attitude measurement system that can provide real-time highprecision attitude feedback information for the control of the boarding bridge telescopic arm,so as to improve the safety of operation and maintenance.First,the boarding bridge telescopic arm attitude measurement hardware system was built,using Jetson AGX Xavier as the processing core,JY61 P IMU to measure the motion information of the telescopic arm,and a USB camera to capture the wind turbine logo image;the YOLOv5-based Jetson AGX Xavier development configuration was completed,and Camera Calibrator toolbox was used to complete the calibration of the camera.Secondly,a vision-based attitude measurement method of the boarding bridge telescopic arm is designed,which uses the YOLOv5 s wind turbine logo detection model to detect the wind turbine logo,and the attitude angle of the boarding bridge telescopic arm is estimated by the EPn P algorithm based on the detection results.Based on the basic principles of YOLOv5 s algorithm and EPn P algorithm,a vision-based algorithm for telescopic arm attitude measurement of boarding bridges was designed and implemented in Jetson AGX Xavier.Test experiments were conducted with the help of a rotary table,and the experimental results showed that the m AP of the target detection model for wind turbine logo detection could reach 96.2%,the attitude angle output frequency of the visual attitude measurement system was 25 Hz,and the static measurement error was less than 0.7°,but the dynamic performance still needed further improvement.Thirdly,we study the basic principle of JY61 P attitude solution,implement JY61 P IMU access and JY attitude solution algorithm in Jetson AGX Xavier,and test the JY attitude solution system with the help of turntable.The experimental results show that the attitude angle output from JY attitude solution system has high instantaneous accuracy and data update frequency up to 100 Hz,but there is a growing drift error,therefore,other attitude measurement systems can be introduced to correct the JY attitude solution system.Fourth,designed a vision/IMU-based attitude data fusion scheme,a state-space model of the boarding bridge telescopic arm attitude measurement system is derived,and an anti-disturbance EKF algorithm is proposed,which corrects the measurement information by dividing the measurement values into three cases: credible,basic credible and uncredible,and improves the anti-disturbance performance of the EKF filtering algorithm.The experimental results show that the combined vision/IMU-based boarding bridge telescopic arm attitude measurement system improves the attitude measurement accuracy by 40.37% over the visual attitude estimation system,the dynamic performance is significantly improved,and the data update frequency can reach 100 Hz.The vision/IMU-based attitude measurement scheme and system of the boarding bridge telescopic arm proposed in this paper can output real-time and high-precision attitude measurement data,which lays the foundation for the implementation of the boarding bridge telescopic arm attitude control system.