Abstract
In this report, the equivalent expression for the mean value of arbitrary non-linear type state variable with respect to the arbitrary non-Gaussian random phenomena is first derived theoretically by use of the expectation operation based on only the standard Gaussian distribution function. Next, based on this convenient expression, an estimation method of the noise evaluation index, Le, is newly derived in a hierarchical form containing the well-known usual estimation method derived under the assumption of the standard Gaussian distribution as the first expansion term. That is, this method is generally applicable to any kind of random phenomena by grasping the statistical information on the noise level fluctuation. Finally, the validity of the proposed method is experimentally confirmed too by applying it to the actual road traffic noise data.
