Bias Errors in a Random Vibration Extremal Control Strategy

摘要

The paper discusses the theoretical basis for an extremal control strategy in random vibration testing. It is shown that except for special cases this strategy results in bias errors. Formulas for estimating the bias are given, and methods for minimizing the errors are suggested. It is concluded that a sequential averager should not be used for this control strategy as its use will result in an unknown bias error. (ERA citation 05:001617)