针对煤矿运输路线复杂，巷道内岔路口和弯道较多的现状，易给井下运输安全带来隐患；设计了一种煤矿井下无轨胶轮车行车岔路口和弯道识别方法。首先，将中线点结构化，形成巷道地图；其次，利用相邻时刻的车辆位置关系，确定行进方向；再次，结合巷道地图结构属性，判定车辆前方是否存在岔路口或弯道；然后，采用迪杰斯特拉（Dijkstra）算法，求解车辆当前位置至疑似目标的最短路径，进而判定是否为对向目标；最后，借助车载终端为司机提供安全性语音提示。在某矿的试验结果表明：方法逻辑判断准确，逻辑响应时间约为32 ms。

In view of the complex transportation route of coal mine, there are many forks and corners in the roadway, which is easy to bring hidden dangers to the safety of underground transportation. In this paper, a method for identifying the intersection and curve of trackless rubber-tyred vehicle in coal mine is designed. Firstly, the method structured the midline points to form a roadway map. Secondly, the direction of travel is determined by using the vehicle position relationship at adjacent moments. Thirdly, combined with the structural attributes of the roadway map, it is determined whether there is a fork or curve in front of the vehicle. Then, the Dijkstra algorithm is used to solve the shortest path from the current position of the vehicle to the suspected target, and then determine whether it is the opposite target. Finally, with the help of the vehicle terminal, the driver is provided with a safe voice prompt. The method is tested in a mine. The results show that the functions are running normally, and the logical response time is about 32 ms.