基于随机森林的群控立体车库运行模式识别研究  

Stereoscopic Garage Group Control Traffic Pattern Recognition Based on Random Forests

作　　者：刘美菊[1] 彭堰艳[1] 龚志恒[1] 刘剑[1] 刘冬[1] 

机构地区：[1]沈阳建筑大学信息控制工程学院,沈阳110168

出　　处：《电子器件》2014年第3期514-518,共5页Chinese Journal of Electron Devices

基　　金：住房与城乡建设部基金项目(2010-K9-22);沈阳建筑大学科研课题项目(2012007)

摘　　要：运用随机森林算法设计出一种立体车库运行模式识别方法。通过采集当前升降机运行中输入的有效数据,组成原始训练数据,利用随机森林算法建立运行模式识别模型,代入立体车库的运行参数后,得到群控立体车库的运行模式。实验结果表明,随机森林算法在立体车库运行模式识别中,能有效识别出当前运行模式。该算法适用于识别立体车库群控系统的运行模式,具有广泛的扩展性。The Random Forest Algorithm is adopted to design a stereoscopic garage pattern recognition system. By collecting the valid data in the current lifts operation to form the primary training data, it establishes a pattern recognition model by using the Random Forest Algorithm, and then it brings the operating parameters of the lifts into the pattern recognition model which is built on the basis of the Random Forest Algorithm to obtain the traffic mode of the group control stereoscopic garage. Experimental results show that the random forests algorithm can effectively identify the current traffic patterns. The experimental results explain that the algorithm is applicable to identify the stereoscopic garage modes of group control system and it has the extensive scalability.

关 键 词：随机森林 群控 立体车库 模式识别 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]