列车轮对状态的融合监测系统

为了提高列车轮对故障诊断准确率和改善现有列车轮对状态在线监测方法的不确定性,结合多传感器信息融合原理,设计了列车轮对融合监测系统,采用特征层融合自适应加权算法进行了轮对状态融合监测,以自适应的方式寻求最优加权因子,使状态测量值总均方误差最小,比较了特征层融合自适应加权算法、模糊数据关联算法、变结构多模的状态估计算法和BP神经网络算法的计算结果。比较结果表明:当轮对两端轴承均出现故障后,两传感器输出的测量值分别为22.0470和21.0250,而此融合算法计算出的估计值为4.2642,融合值最接近真值,因此,列车轮对融合监测系统可靠性高,抗干扰性强。

Fusion monitoring system of locomotive wheelset state

In order to improve the accuracy of fault diagnosis and the uncertainty of current online condition monitoring methods for locomotive wheelset,a fusion monitoring system of locomotive wheelset was designed based on multi-sensor information fusion principle. The state of locomotive wheelset was monitored by using feature level fusion adaptive weighting algorithm,and the measured values were weighted adaptively to obtain the least-mean-square error of the measured values. The results computed by feature level fusion adaptive weighting algorithm,fuzzy data association algorithm,variable structure multiple-model estimation algorithm and BP nerve network(BPNN) algorithm were compared. Comparison result shows that when the fault occurs in the bearings of wheelset,the measured values are 22.047 0 and 21.025 0 respectively,while the estimation value from the fusion algorithm is 4.264 2,so the system has high reliability and better anti-disturbance. 1 tab,6 figs,13 refs.