为解决轴承信号采集时难以确定最优测点这一问题。以NU306轴承-轴承座系统为研究对象，对轴承座开展多测点振动测试实验，运用局部线性嵌入算法对轴承多维数据进行线性降维得到测点灵敏度矩阵，研究了内、外圈故障时，载荷与转速变化下最优测点位置。结果表明，对于外圈故障的轴承，转速或载荷增大时优质测点为距离故障位置最近的测点；对于内圈故障的轴承，载荷或转速增大时优质测点为距离轴承座中心远且载荷较小的测点。研究结果可为不同工况下的测点选择提供有效参考。

In order to solve the problem of difficulty in determining the optimal measurement points for bearing signal acquisition. Taking the NU306 bearing-housing system as the research object, the bearing housing to carry out multi-measurement points vibration test experiment. The local linear embedding algorithm was used to linearly reduce the multi-dimensional space data of the bearings to obtain the sensitivity matrix of the measurement point. The optimal position of the measuring point under varying load and rotation speed was studied when the inner or outer ring are defective. The results show that for bearings with outer ring failures, the optimal measurement point is the one closest to the location of the failure when the load or rotational speed increases. For bearings with inner ring failures, the optimal measurement point is the one at lower loads far from the center of the housing when the load or rotational speed increases. The results can provide an effective reference for the selection of measurement points under different operating conditions.