Editing MISTRAM (section)

==Minuteman Inertial Guidance System testing==
The first [[LGM-30 Minuteman|Minuteman missiles]] (MM I) were launched in the early 1960s from the Air Force Eastern Test Range (AFETR) and were tracked with the AZUSA CW tracking system. The comparatively low quality of the AZUSA tracking data, combined with the rudimentary stage of evaluation techniques, allowed only estimation of the total error; no isolation of individual inertial measurement unit (IMU) error sources was possible.<ref name="aiaa">{{cite conference|author1=R. Fuessel |author2=J. McGhee |author3=R. Powers |author4=D. Sifter  |name-list-style=amp |title=A method for determining the performance of a precision inertial guidance system|book-title=AIAA Guidance and Control Conference, August 6–8, 1979, Boulder, Colorado. AIAA Paper No. 1979-1891|pages=637–644}}</ref>

Subsequent development of improved tracking systems, UDOP and MISTRAM, at AFETR yielded much higher quality velocity tracking profiles. During the Minuteman II flight test program, significant improvements were made in the post-flight evaluation of the IMU accuracy. The most important of these improvement was the introduction of maximum likelihood error estimation using the Kalman algorithm to filter the velocity error profile. Continued improvement of the UDOP and MISTRAM tracking systems and refinement of the evaluation techniques during the Minuteman III flight test program made it possible to gain considerable insight into NS-20A1 IMU error sources.<ref name="aiaa"/>

===Accuracy evaluation===
One of the major problems in trajectory and orbital estimation is to obtain a realistic estimate of the accuracy of the trajectory and other important parameters. In the orbital case, some of the parameters which may not be solved for are geopotential constants, survey, etc. These factors will affect the total uncertainty in the orbit and, of course, ephemeris predictions. A statistical technique was developed that performs a variance-covariance propagation to obtain accuracy estimates based on random and unmodeled errors. An example of the unmodeled error propagation in the MISTRAM system was given for the Geos B satellite.<ref name="technometrics">{{cite journal|author=Norman Bush|title=Unmodeled Error Analysis on Trajectory and Orbital Estimation|journal=Technometrics|volume=13|issue=2|date=May 1971|pages=303–314|doi=10.2307/1266792|jstor=1266792}}</ref>