Abstract:

In the environmental vibration test of a structural part， it is necessary to use multiple sensors to control the excitation applied to the test sample， so that the power spectrum input to the test sample would be basically consistent with the reference spectrum. However， the power spectrum measured by each of the control sensors may be significantly different， resulting in a noticeable inconsistency between the magnitudes of vibration forces transferred to the test sample by fixture. In some cases， the control spectrum may even go out-of-tolerance to interrupt the test. To address such a problem， this paper carries out the research on the optimization design of the control sensor positions for a vibration test. Firstly， the mathematical model of the position optimization of the control sensors is established according to the multi-point control strategy of the random vibration test on the vibration table. Using the feature mapping method， we obtain the frequency response function of a control sensor by the weighted sum of the frequency response function at the nodes around the control point to ensure continuous movement of the control sensor during the optimal design process. Then， using the gradient-based optimization algorithm， we can readily achieve the optimal position design of the control sensors， and considerably reduce the difference between the root-mean-square value of the response spectra measured by the control sensors. The optimization method proposed in this paper can provide a theoretical guidance for the position design of control sensors in the environmental vibration test of a structural part.

Key words: sensor layout optimization, feature mapping, power spectral density, sensitivity analysis, excitation function fidelity