Apr 1, 2000
Optimal Choice of Track Fitting Procedure for Contaminated
Data in High-Accuracy Cathode Strip Chambers
Joint Institute for Nuclear Research, Dubna
Modern detectors of elementary particles achieve a very high
accuracy of coordinate measurements (the order of tens microns)
However heavy physical backgrounds can decrease the accuracy in
several orders of magnitude. Therefore one needs to choose an
optimal track fitting procedure, i.e. a simple calculating
approach guaranteeing the best track parameters estimation.
This task is especially important for processing of contaminated
data from high-accuracy cathode strip chambers. Since the
traditional least squares method (LSQ) looses its optimal
properties on contaminated data, some of authors made attempts
to apply an LSQ modification with refitting track after rejecting
more distant points. A "straight" application of maximum
likelihood method (MLM) improves the situation significantly,
but leads to quite cumbersome calculating procedures.
Thus the problem is to choose a simple calculating procedure
giving the most accurate result. We propose a robust track fitting
procedure with sub-optimal weight function. Our approach can be
characterized by a mathematical simplicity, easy-calculated weights
and a high-speed program realization.
In our comparative study we estimate track parameters on data simulated
for CMS endcap muon chambers applying a pure LSQ method, LSQ method
with point rejecting, a "straight" MLM and our robust technique.
The results obtained demonstrate the advantages of our robust approach.
|Short Paper: |