A Novel Credibility Quantification Method for Welch's Periodogram Analysis Result in Model Validation
Yuchen Zhou, Ke Fang, Kaibin Zhao, Ping Ma · Linköping electronic conference proceedings · 2018
Welch's periodogram is widely used in frequency domain model validation.However, Welch's analysis results just reveals whether the time series passed the consistency test in each discrete frequency point, which is not a quantitative credibility evaluation result and may not help the evaluation expert to grade credibility level of simulation system.Based on Welch's periodogram and consistency test approach, a novel credibility quantification method using weight density function is proposed.Furthermore, the frequency analysis and credibility quantification process is provided.Finally, the credibility quantification of radiated noise in ship acoustic feature simulation indicates the method proposed is effective for periodic time series with complicated spectrum.