Optimal tuner selection for Kalman filter-based aircraft engine performance estimation

Original publisher: Cleveland, Ohio : National Aeronautics and Space Administration, Glenn Research Center, [2010] OCLC Number: (OCoLC)681412392 Excerpt: ... Estimate Estimate Subset of health parameters Subset of health parameters 10 10 Actual Actual % Fn % Sm 0 0 Δ Δ − 10 − 10 0 50 100 150 200 250 300 0 50 100 150 200 250 300 SVD tuner selection SVD tuner selection 10 10 % Fn % Sm 0 0 Δ Δ − 10 − 10 0 50 100 150 200 250 300 0 50 100 150 200 250 300 Systematic tuner selection Systematic tuner selection 10 10 % Fn % Sm 0 0 Δ Δ − 10 − 10 0 50 100 150 200 250 300 0 50 100 150 200 250 300 time ( sec ) time ( sec ) Figure 5. - SmLPC estimation ( tuner comparison ). Figure 4. - Fn estimation ( tuner comparison ). TABLE 5. - HEALTH PARAMETER PERCENT SQUARED ESTIMATION ERRORS ( NOMINAL NOISE ) Estimator Error type h1 h2 h3 h4 h5 h6 h7 h8 h9 h10 Sum Theor. squared bias 1.26 1.48 2.28 1.22 0.00 0.00 0.74 0.00 1.97 3.06 12.00 Theor. variance 0.73 0.19 0.80 0.45 0.40 0.68 0.21 0.21 0.34 0.08 4.09 Kalman filter Theor. squared error 2.00 1.67 3.07 1.66 0.40 0.68 0.96 0.21 2.32 3.14 16.09 Exper. squared error 1.82 1.60 2.77 1.59 0.41 0.68 0.97 0.21 2.13 3.15 15.31 Theor. squared bias 2.42 1.75 3.53 1.90 0.47 0.96 1.00 0.16 2.47 3.17 17.81 Theor. variance 0.22 0.27 0.16 0.64 0.69 0.71 0.36 0.36 0.17 0.09 3.68 MAP estimator Theor. squared error 2.63 2.02 3.69 2.54 1.16 1.67 1.36 0.52 2.64 3.26 21.48 Exper. squared error 2.52 1.89 3.44 2.45 1.14 1.57 1.34 0.51 2.49 3.24 20.60 TABLE 6. - HEALTH PARAMETER PERCENT SQUARED ESTIMATION ERRORS ( REDUCED NOISE ) Estimator Error type h1 h2 h3 h4 h5 h6 h7 h8 h9 h10 Sum Theor. squared bias 1.26 1.48 2.28 1.22 0.00 0.00 0.74 0.00 1.97 3.06 12.00 Theor. variance 0.02 0.00 0.03 0.01 0.01 0.02 0.00 0.00 0.01 0.00 0.10 Kalman filter Theor. squared error 1.29 1.48 2.30 1.22 0.01 0.02 0.74 0.00 1.98 3.06 12.10 Exper. squared error 1.13 1.41 1.97 1.16 0.01 0.02 0.75 0.00 1.79 3.08 11.31 Theor. squared bias 1.26 1.48 2.28 1.22 0.00 0.00 0.74 0.00 ...