### ASA 129th Meeting - Washington, DC - 1995 May 30 .. Jun 06

## 2pNS3. Statistical methods for automated INM path derivation.

**David A. Flynn
**

**
**
*Larson Davis Labs., 1681 W 820 N, Provo, UT 84601
*

*
*
Flight paths at an airport can be viewed as samples of a statistical
distribution. An individual sample path x[sub i], is described by the position
as a function of distance from the airport. The mean paths of distribution
minimize (Sigma)x[sub i]-, where the operator ``-''
is an approximation of the difference in ground noise projection of two paths,
and ``m'' is the desired number of mean paths. Cluster analysis methods can be
used to find this set of mean paths: they will by definition be the optimal m
paths for use in the INM in place of the original sample paths. These methods
can reduce hundreds of thousands of radar monitored paths into a concise set of
INM tracks. Moreover, they do not suffer measurement anomalies that gate
penetration analysis suffers.