多传感器数据融合在爬架控制系统中的应用  

Application of multi-sensor data fusion algorithm in control system of climbing scaffold

作　　者：陈斌[1] 金堃[1] 李宏[1] 

机构地区：[1]宁波大学信息科学与工程学院,浙江宁波315000

出　　处：《传感器与微系统》2015年第9期151-153,157,共4页Transducer and Microsystem Technologies

基　　金：浙江省教育厅科研资助项目(Y201329962);宁波大学教研项目(JYXMXSJ201406)

摘　　要：针对传统爬架控制系统存在安全性评估不全面、报警准确率低的缺陷,提出了一种基于多传感器数据融合技术以提高爬架控制系统安全预警性能的方法。该方法通过传感器采集爬架载荷、倾斜角度和爬升速度等数据,利用BP神经网络对数据进行融合,并对系统的安全性能做出评估。详细介绍了融合算法、训练过程、系统硬件构成和软件实现。测试结果表明:该方法不仅能多角度评估系统的安全性,而且能提高爬架控制系统报警准确性,从而提高爬架系统的可靠性。Aimed at disadvantages of incomprehensive safety assessment and low alarm accuracy in traditional climbing scaffold control system,put forward a method to improve safety warning performance of climbing scaffold control system by multi-sensor data fusion technology. The method acquires data of load,oblique angle and rising velocity of climbing scaffold through sensors,fuse data by using BP neural network and then evaluate safety of system. Describe fusion algorithm in detail,training process,system hardware and software realization. The test results show that the method not only can assess security of system from multi-angle,but also can improve alarm accuracy of climbing scaffold control system,so as to improve reliability of the system.

关 键 词：爬架 数据融合 神经网络 控制系统 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]