Performance analysis of nonlinear filters using dynamic error spectrum metric

The commonly used root-mean-square error (RMSE) for estimation performance evaluation is easily dominated by large error terms. Then many new alternative absolute metrics has been provided. But each of these metrics only reflects one narrow aspect of estimation performance respectively. A comprehensive measure, error spectrum, was presented aggregating these incomprehensive measures. However, when applying this measure to dynamic systems, it will plot a 3D figure over the total time span, which is not intuitive and difficult to be analyzed. In this study, to overcome its drawbacks, the authors propose a new metric, dynamic error spectrum, to summarize the ES curve. Three forms under different application backgrounds are given, one of which is balanced taking into account both good and bad behavior of an estimator and so can provide more impartial evaluation results. It can be applied to a variety of dynamic systems directly. Then considering the challenge in performance evaluation of nonlinear filters for nonlinear system, we choose four nonlinear filters to illustrate the superiority of our metric. The simulation results validate its utility and effectiveness.