摘要
地面传感器阵列对目标的分类能力是影响其使用效能的关键因素之一;传感器阵列获取的观测数据受外部环境影响,当外部各种物理场特征随时间发生改变时,采用固定的分类标准进行判断会降低传感器阵列对目标分类识别结果的可信性;提出基于动态数据驱动的传感器阵列目标分类方法,对时变的外部环境特征加以提取作为目标识别计算的反馈控制量,构建环境改进模型作为决策用有限状态自动识别机,为其中每个状态训练相应的分类方法,实现对分类准则的自适应选择,提高算法性能。
The classification ability of the ground sensor array is one of the key factors that affect its use efficiency.Sensor array to obtain the observation data are influenced by the external environment,when an external characteristics change over time,various physical fields using fixed classification standard to judge will reduce the credibility of the sensor array of target classification recognition results.Proposed sensor array object classification method based on dynamic data driven,the external environment of time-varying feature extraction for target recognition calculation amount of feedback control,build environment improved model as decision with finite state recognition machine automatically,for each of these state training corresponding classification method,the adaptive selection of classification standards,improve the algorithm performance.
作者
永胜
Yong Sheng(Finance Department, Inner Mongolia Autonomous Region,Hohhot 010098,Chin)
出处
《计算机测量与控制》
2018年第5期278-281,共4页
Computer Measurement &Control