摘要
由于重力传感器容易受环境温度等各种非目标参量的影响,其输出性能大大降低,为此,采用Elman神经网络多传感器融合技术加以解决。传感器信息融合是对多种信息的获取、表示及其内在联系进行综合处理和优化的技术。多传感器信息经过融合后能够完善、准确地反映环境的特征。采用Elman神经网络补偿由系统的非线性和外界干扰引起的误差。仿真试验表明,该算法计算量小、拟合程度好、精确度高。
While the output performance of the gravity sensors is greatly deteriorated because their device characteristics are easily affected by environment temperature and various non-target parameters, Elman neural network multi-sensor fusion technology is adopted to solve the problem. Sensor infomlation fusion is the technology that is able to comprehensively process and optimize variety information acquisition, representation and their intrinsic relations. With fusion, the information of multi-sensors can precisely and perfectly reflect the characteristics of environment. The errors caused by the nonlinearity of system and external interference are compensated by adopting Elman neural network. The simulation indicates that the algorithm features small amount of calculation, goodness of fitting, and high accuracy.
出处
《自动化仪表》
CAS
北大核心
2012年第11期66-68,71,共4页
Process Automation Instrumentation
关键词
传感器
数据融合
神经网络
动态补偿
鲁棒性
Sensor Data fusion Neural network Dynamic compensation Robustness
