Research On Road Condition Recognition System By Multi-sensor Information Fusion For All-terrain Crane

Off-highway vehicles undertake the tasks of emergency rescue and engineering construction,and their efficiency,safety and driving comfort affect the efficiency and safety of construction.This article takes the all-terrain crane as the research object,combined with the common road conditions when the all-terrain crane is in transit,studies a set of systems that can identify the road surface static obstacles and road surface level information,referred to as road condition recognition system.The road condition recognition system includes a hardware system and a software system: the hardware system uses three-dimensional lidar as the main sensor for road condition information recognition.In addition,the photoelectric encoder cannot accurately measure the odometer when the wheeled off-highway vehicle slips,and can only obtain a single In the case of a DOF odometer,a stereo camera and an inertial measurement unit are fused to calculate a six-DOF odometer,and finally all sensors are installed on the designed bracket;the software system is based on the open source robot operating system Robot Operating System as the framework,using open source points The cloud processing function library Point Cloud Library processes the real-time road data of the three-dimensional lidar.Based on the Kalman filter principle,the visual odometer calculated by the stereo camera and the inertial odometer of the inertial measurement unit are calculated to obtain a six-degree-of-freedom visual inerometer.The odometer corrects the attitude of the point cloud,eliminates the error of the point cloud,and superimposes the point cloud based on the recursive idea.Finally,the corrected point cloud road condition information is used to identify the road surface static obstacles,and the road surface grade is evaluated with reference to the national standard for road surface grade evaluationThe road condition recognition system was installed on a five-bridge all-terrain crane of a certain model for real vehicle experiments,and high-precision single-point laser measurement and high-precision vibration sensors were used to measure the elevation information of the same pavement,which was used as the real elevation of the pavement for the road condition recognition system The results of the identification were verified.The test site and road surface were a company’s closed vehicle comprehensive test site.After the experiment,the road condition recognition data was analyzed and sorted,and the effectiveness of the designed road conditionrecognition system was evaluated based on the data results,and suggestions were made to improve the road condition recognition system.