基于虚拟墙的汽车起重机吊装防碰撞技术

发布时间：2018-11-03 19:14

【摘要】：为减少汽车起重机在吊装作业过程中吊臂与作业区障碍物发生碰撞的事故,建立障碍物分类模型,采用吊臂臂头自主探测学习方法,准确采集作业区障碍物的外形与位置信息。建立起重机圆柱坐标系,作业区按回转角度分成不同扇区,扇区按距回转中心距离分成不同扇格,以扇格精确存贮作业区环境数据,利用扇格数据生成吊装安全作业三维虚拟墙。以吊臂最近的20个时刻位置信息,用加权线性回归模型预测下一时刻吊臂的动作与位置参数。针对吊臂的预测动作及与虚拟墙的不同距离,采取不干预、限速、微动与禁止等吊臂控制策略。实际作业环境下的功能测试表明,该系统能有效地防止吊臂与作业区障碍物发生碰撞。
[Abstract]:In order to reduce the accident of collision between the boom and the obstacle in the working area during the hoisting operation, the classification model of the obstacle is established, and the information of the shape and position of the obstacle in the operation area is accurately collected by the self-detection and learning method of the jib head. The cylindrical coordinate system of crane is established. The operation area is divided into different sectors according to the rotation angle, and the sector is divided into different fan lattices according to the distance from the center of rotation, so as to store the environmental data of the operation area accurately. The three-dimensional virtual wall of lifting safety operation is generated by fan-lattice data. Based on the position information of the most recent 20 times of the boom, the weighted linear regression model is used to predict the movement and position parameters of the next moment. Aiming at the predicted action of the boom and the different distance from the virtual wall, the control strategies of non-intervention, speed limit, fretting and forbidding are adopted. The functional test in the actual working environment shows that the system can effectively prevent the collision between the boom and the obstacles in the operation area.
【作者单位】：长安大学信息工程学院;徐州工程机械集团有限公司;
【基金】：中国博士后科学基金面上(2013M532003) 江苏省博士后集聚计划(2012097) 中央高校基本科研业务费专项资金(CHD2011JC179)资助项目
【分类号】：TH213.6

【参考文献】