中文摘要：

针对封锁雷在无人值守状态下如何有效防止敌方排雷动作，延长自身对目标区域的封锁控制时间的问题，采用三轴 MEMS 加速度传感器构建一种新型的封锁雷智能防排系统。设计试验采集封锁雷静止姿态和运动状态特征信号，采用 SMA 和 DSVM 算法对特征信号进行分析，提取特征参数，设计二叉决策树特征分类器，利用特征信号对分类器进行参数训练。试验结果表明，建立的封锁雷智能防排系统针对5种静止姿态和3种运动状态具有较好的识别率，所建系统有效可行。

英文摘要：

In order to prevent the blockage mine from being eliminated by the enemy in the unattended operation state and to postpone the work time in the target area,a novel intelligent anti removal system for blockage mines based on a tri-axis MEMS accelerometer was established. An experiment was designed to acquire the characteristic signal of the blockage mine including the static posture and the dynamic state. The SMA and DSVM algorithms were introduced to analyze the characteristic signal and abstract the character parameters. A binary decision tree charac-ter classifier was designed based on these parameters and the parameter of the classifier was trained from the charac-teristic signal. The verification experiment results show that the recognition accuracy of the blockage mine′s five static postures and three dynamic processes is relatively high and as a result the established system is effective and feasible.