CR-Calculus and adaptive array theory applied to MIMO random vibration control tests

13th International Conference on Motion and Vibration Control; 12th International Conference on Recent Advances in Structural Dynamics

Musella, U.^1,2 ; Manzato, S.^1 ; Peeters, B.^1 ; Guillaume, P.^1

Siemens Industry Software NV, RTD Test Division, Interleuvenlaan 68, Leuven

3000, Belgium^1

Vrije Universiteit Brussel, Mechanical Engineering Department, Pleinlaan 2, Elsene

1050, Belgium^2

关键词: Performance comparison;    Phase information;    Process performance;    Realistic environments;    Spectral density matrix;    Time waveform replications;    Vibration control strategies;    Vibration environment;   

Others  :  https://iopscience.iop.org/article/10.1088/1742-6596/744/1/012175/pdf
DOI  :  10.1088/1742-6596/744/1/012175

来源: IOP

【 摘 要 】

Performing Multiple-Input Multiple-Output (MIMO) tests to reproduce the vibration environment in a user-defined number of control points of a unit under test is necessary in applications where a realistic environment replication has to be achieved. MIMO tests require vibration control strategies to calculate the required drive signal vector that gives an acceptable replication of the target. This target is a (complex) vector with magnitude and phase information at the control points for MIMO Sine Control tests while in MIMO Random Control tests, in the most general case, the target is a complete spectral density matrix. The idea behind this work is to tailor a MIMO random vibration control approach that can be generalized to other MIMO tests, e.g. MIMO Sine and MIMO Time Waveform Replication. In this work the approach is to use gradient-based procedures over the complex space, applying the so called CR-Calculus and the adaptive array theory. With this approach it is possible to better control the process performances allowing the step-by-step Jacobian Matrix update. The theoretical bases behind the work are followed by an application of the developed method to a two-exciter two-axis system and by performance comparisons with standard methods.