Using Statistical Analysis to Extract
Meaningful Beam Position Monitor Data
Student: Eric Thompson
School: Christopher Newport University
Mentored By: Rui Li
BPM data taken from rayTrace measurement at CEBAF consist of betatron orbits, BPM noise, residual dispersions, and malfunctioned BPM signals. To extract meaningful betatron orbits from the noise, as well as identify contributions of different noise sources and their behaviors, model independent analysis (MIA) was conducted on the rayTrace data. Singular value decomposition (SVD, a method of statistical analysis) was used to identify principle components of the BPM data. Our results clearly demonstrate dominant betatron orbits, less dominant residual dispersion, signals from malfunctioned BPMs, and BPM noises. Here the residual dispersion appears as the spatial dependence of the horizontal orbit, which is comparable with Elegant generated dispersion based on the design model. SVD also reveals the temporal dependence related to each principle component, such as kicker strength and energy variation. Finally, Convergence checks were performed by varying the number of BPM orbits and number of BPMs used for SVD analysis, and these results are presented here.