Table 2.3: Crista IMU HWIL Sensor noise models PDynamic PStatic G unit P Pa] [deg/s] [m/s/s) Resolution unit 3.906 200|1.6E-46.0E-3 Min unit 300 0.0-5.20|-1000 4000110.0005.20 100.0 20.0 Butterworth Order 2 2 2 2 BW Cutoff Freq. Hzl 11.0 11020.0 20.0 Drift Rate unit 0.05 1.0 1.5E-42.0E-3 Max Drift value [unit] 15.0 100.00.01 0.20 Drift Update Rate [ s 5.0 5.0 1.0 1.0 2.1.4 Dryden Turbulence Stochastic turbulence disturbances are required for accurate HWIL simulations, as eal world experiments are characterized by unpredictable winds acting on the vehicle The Dryden turbulence model is one of the accepted methods for including turbulence in aircraft simulations 28. By applying shaped noise with known spectral properties as velocity and angle rate perturbations to the body axes of the vehicle, the effect of turbulence is captured during discrete time simulations. The noise spectrum for each of the perturbations is predominantly described by a turbulence scale length parameter L e airspeed reference velocity, Vo, and the turbulence intensity,o The selection of these parameters allows for the turbulence to be modeled accordin to the prevailing wind conditions The spectral frequency content for generalized aircraft turbulence have been well studied (29, 28 and are given for each of the aircraft body axes r1+(u)2 Suq(w) L。1+3(÷)2 (211) +( n()=9L.1+3({) (1+(÷=u)2Table 2.3: Crista IMU HWIL Sensor Noise Models Sensor PDynamic PStatic Gyro Accel. [unit] [Pa] [Pa] [deg/s] [m/s/s] Resolution [unit] Min [unit] Max [unit] Noise Gain Butterworth Order BW Cutoff Freq. [Hz] Drift Rate [unit/s] Max Drift value [unit] Drift Update Rate [s] 3.906 ­300 4000 20.0 2 11.0 0.05 15.0 5.0 2.00 0.0 110,000 20.0 2 11.0 1.0 100.0 5.0 1.6E­4 6.0E­3 ­5.20 ­100.0 5.20 100.0 0.10 0.0 2 2 20.0 20.0 1.5E­4 2.0E­3 0.01 0.20 1.0 1.0 2.1.4 Dryden Turbulence Stochastic turbulence disturbances are required for accurate HWIL simulations, as real world experiments are characterized by unpredictable winds acting on the vehicle. The Dryden turbulence model is one of the accepted methods for including turbulence in aircraft simulations [28]. By applying shaped noise with known spectral properties as velocity and angle rate perturbations to the body axes of the vehicle, the effect of turbulence is captured during discrete time simulations. The noise spectrum for each of the perturbations is predominantly described by a turbulence scale length parameter, L, the airspeed reference velocity, Vo, and the turbulence intensity, σ. The selection of these parameters allows for the turbulence to be modeled according to the prevailing wind conditions. The spectral frequency content for generalized aircraft turbulence have been well studied [29, 28] and are given for each of the aircraft body axes: Sug(ω)= 2σ2 uLu πVo · 1 1 + ( Lu Vo ω)2 (2.10) Svg(ω)= σ2 vLv πVo · 1 + 3( Lv Vo ω)2 � 1 + ( Lv Vo ω)2 �2 (2.11) Swg(ω)= σ2 wLw πVo · 1 + 3( Lw Vo ω)2 � 1 + ( Lw Vo ω)2 �2 (2.12) 37