改进双向A^(*)算法的矿井火灾逃生实时动态路径规划研究

Research on Real-time Dynamic Path Planning for Mine Fire Escape with Improved Bidirectional A^(*) Algorithm

针对煤矿井下工作环境的复杂性和危险性,当发生矿井火灾时,为了提高井下人员的生存率,提出了一种改进双向A^(*)算法的矿井火灾逃生实时动态路径规划方法。借助井下布设的传感器监测设备获取的实时监测数据,动态计算权值,对双向A^(*)算法的评估函数进行改进。综合考虑了巷道类型、坡度、障碍物、风速风向4种影响井下人员逃生的因素,基于国内某煤矿采集的部分真实井下巷道数据,结合实时的传感器监测数据,对算法进行验证并与Dijkstra算法、传统A^(*)算法和双向A^(*)算法进行对比,试验结果表明:改进后的双向A^(*)算法不仅能够有效减少存储成本,而且在不同的条件下都能够动态规划出路线节点更少、最优的安全逃生路线,有助于提高作业人员逃生效率,降低人员伤亡率。

In response to the complexity and danger of the working environment in coal mines,a real-time dynamic path planning method for mine fire evacuation based on an improved bidirectional A^(*)algorithm was proposed to improve the survival rate of underground personnel in the event of a mine fire.By utilizing real-time monitoring data obtained from sensor monitoring equipment deployed underground,dynamically calculating weights,and improving the evaluation function of the bidirectional A^(*)algorithm.Four factors were took into account that affect the escape of underground personnel,including tunnel type,slope,obstacles,wind speed and direction,and based on some real underground tunnel data collected from a domestic coal mine,combined with real-time sensor monitoring data,the algorithm was validated and compared with Dijkstra algorithm,traditional A^(*)algorithm,and bidirectional A^(*)algorithm.The experimental results showed that the improved bidirectional A^(*)algorithm could not only effectively reduce storage costs,but also under different conditions,it is possible to dynamically plan the optimal safe escape route with fewer route nodes,which helps to improve the evacuation efficiency of workers and reduce the rate of casualties.