CAJ | 학술논문

研究发现，通过低信噪比的实测数据得到的单一关联维数识别电梯的运行状态，是不全面的，易出现误判。这里提出以多维度关联维数联合判据作为识别依据，并以最小二乘支持向量机为识别方法。结果表明：多维度关联维数联合判据能够反映更加丰富的电梯动力学特性，识别准确度较高。
연구발현，통과저신조비적실측수거득도적단일관련유수식별전제적운행상태，시불전면적，역출현오판。저리제출이다유도관련유수연합판거작위식별의거，병이최소이승지지향량궤위식별방법。결과표명：다유도관련유수연합판거능구반영경가봉부적전제동역학특성，식별준학도교고。
The study found that using the single correlation dimension calculated from the vibration signal of elevator which is low signal-to-noise ratio, recognizes the running states of elevator. The resolution is not comprehensive enough and prone to miscarriage of justice. In this paper,we make the combined criterion of correlation dimension to multiple dimensions as recognition basis,and with the least square support vector machine method. The results show that the combined criterion reflects dynamic characteristic of elevator richer,and the recognition accuracy is relatively high.