微型电脑应用

现有巡检机器人检测方法受到机器人各参数量纲不同的影响，异常报警效果差，经常出现漏检和误检问题，对此，设计一个不间断巡检机器人的仓库异常自主报警方法。采用蚁群算法规划机器人运动路径，实时更新巡检路线，保证机器人不间断巡检；通过通信协议向机器人控制器发送约定指令，获取传感器采集的检测数据并进行归一化处理；结合形态学与边缘区域包围度抑制方式识别检测目标，依据重建误差表示每个位置异常的可能性，设定异常情况报警过程，实现巡检机器人的仓库异常自主报警。实验结果表明，所研究的机器人仓库异常自主报警方法的漏检次数、误报次数与误检次数均不超过一次，提高了异常报警效果，具备实际应用意义。

The existing inspection method of inspection robot is affected by the different parameter scale, the abnormal alarm effect is poor, and leak detection and false inspection problems often occur. In this regard, an abnormal automatic alarm method of uninterrupted inspection robot is designed. Ant algorithm is used to plan the robot movement path, update the inspection route in real time to send the agreed instructions to the robot controller to obtain the detection data and normalize it. It indicates the possibility of each position according to the reconstruction error, sets the abnormal situation alarm process to realize the abnormal warehouse alarm of the inspection robot. The experimental results show that the number of missed tests, false positives and false tests of the studied robot warehouse are not more than once, which improves the abnormal alarm effect and has practical application significance.