基于积分链动力学和Gröbner基的露天矿自动驾驶车辆运动规划

Motion planning of open pit autonomous vehicle based on integral chain dynamics and Gröbner basis

  • 摘要: 露天矿自动驾驶能够减少作业人员数量,降低生产成本和安全风险,已成为露天矿智能化发展的重要方向。然而,由于当前自动驾驶算法规划的矿车行驶速度较为保守,导致其运输效率与有人驾驶相比还存在一定差距。为在不违反车路物理约束、保证行驶安全的前提下,尽可能提高规划速度,提出了一种基于积分链动力学和Gröbner基的露天矿既定运输路线运动规划方法。首先,根据矿车运动学特性和露天矿道路情况,综合考虑限速规则、道路曲率、地面附着力以及车辆前轮转角等条件,建立了既定运输路线全路段精细化速度约束曲线;随后,采用积分链动力学理论将露天矿自动驾驶运动规划问题表征为关于急动度作用时间的多元多项式方程组求解问题,利用Gröbner基理论将多元多项式方程组转化为带参数的正则化多项式系统,最终将运动规划问题等价为三角化多项式方程组的求解问题,实际部署中只需输入始末条件参数即可实现快速迭代求解,得到满足驾驶安全需求且加速度平滑的运动规划方案。利用该方法对内蒙古某露天矿真实运输路线的多个路段在不同的初始和终止条件下进行了运动规划,均能快速求得可行解;最后针对3.61 km长的连续弯道往返路线进行了运动规划,与目前该矿有人驾驶矿车的实际运行记录相比,平均运行时间减少10.7%,20 m的规划任务计算时间在100 ms以内,表明该方法能有效提升露天矿车辆的运行效率。

     

    Abstract: Autonomous driving in open-pit mines can reduce the number of operating personnel, lower production costs and safety risks, and has become an important direction for the intelligent development of open-pit mines. However, due to the relatively conservative driving speed of mining trucks planned by current autonomous driving algorithms, there is still a certain gap in its transportation efficiency compared with manned driving. To maximize the planned speed as much as possible on the premise of not violating vehicle-road physical constraints and ensuring driving safety, a motion planning method for fixed transportation routes in open-pit mines based on integral chain dynamics and Gröbner basis is proposed. First, a refined speed constraint curve for the entire section of the fixed transportation route is established according to the kinematic characteristics of mining trucks and road conditions in open-pit mines, with comprehensive consideration of conditions such as speed limit rules, road curvature, ground adhesion, and vehicle front-wheel angle. Subsequently, the integral chain dynamics theory is used to characterize the autonomous driving motion planning problem in open-pit mines as a problem of solving a system of multivariate polynomial equations regarding the action time of force change rate. The Gröbner basis theory is applied to convert the system of multivariate polynomial equations into a parameterized regular polynomial system, and finally the motion planning problem is equivalent to the problem of solving a triangularized polynomial system. In practical deployment, only the start and end condition parameters need to be input to achieve fast iterative solution, thus obtaining a motion planning scheme that meets driving safety requirements and has smooth acceleration. This method is used for motion planning of multiple sections of a real transportation route in an open-pit mine in Inner Mongolia under different initial and terminal conditions, and feasible solutions are quickly obtained in all cases. Finally, motion planning is conducted for a 3.61 km-long continuous curved round-trip route. Compared with the current actual operation records of manned mining trucks in the mine, the average operation time is reduced by 10.7%, and the calculation time for a 20 m planning task is within 100 ms, which indicates that this method can effectively improve the operation efficiency of vehicles in open-pit mines.

     

/

返回文章
返回