Statistical performance analysis simulation of the fast merging procedure fuzzy logic

This paper describes a performance simulation that was developed to evaluate the fuzzy decision logic of the passive Final Approach Spacing Tool (FAST). The problem was to evaluate the performance of the FAST fuzzy logic when the aircraft ground speed input data is corrupted by the radar tracking software. Conventional approaches for evaluating fuzzy logic center around the use of Monte Carlo simulated inputs, multiple computer runs, and computed sample statistics of the fuzzy logic results. In this paper, an error covariance approach was used yielding performance statistics with a single computer run. While FAST uses four different fuzzy logic Procedures, the focus of this paper is on the development of a performance simulation of the FAST Merging Procedure that is used prior to the final flight path segment. The Merging Procedure establishes the preferred sequence of merging aircraft from separate flight path segments onto a common segment in the Terminal Radar Approach Control (TRACON). By modelling both the nomial and statistical performance or this Procedure, Ir is possible to determine if the decision logic might reach the incorrect merging sequence decision due to the ground speed tracking errors. Hence, the nominal, estimated, and the 95% confidence interval decision parameters are determined with this performance simulation. In addition to providing a description of the theoretical basis for this Performance simulation, a simple test case is presented to illustrate the results that can be obtained.