You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: docs/en/advanced/airspeed_validation.md
+50Lines changed: 50 additions & 0 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -117,6 +117,56 @@ To configure the delay before PX4 starts or stops using airspeed sensor data aft
117
117
[ASPD_FS_T_STOP](#aspd_fs_t_stop_table): Delay after failing validation before the sensor is considered invalid.
118
118
119
119
120
+
121
+
## Examples
122
+
123
+
### Flight 1: Blocked Pitot Tube during Rainy Conditions
124
+
125
+
During a fixed-wing flight in rainy conditions, the pitot tube became blocked. The following configuration was used:
126
+
127
+
[ASPD_DO_CHECKS](#aspd_do_checks_table) = 5 ([missing data check](#missing-data-check) and [innovation check](#innovation-check) enabled) \
128
+
[ASPD_FS_INNOV](#aspd_fs_innov_table) = 4 \
129
+
[ASPD_FS_INTEG](#aspd_fs_integ_table) = 10 \
130
+
[FW_AIRSPD_STALL](#fw_aspd_stall_table) = 12 \
131
+
[ASPD_FS_T_STOP](#aspd_fs_t_stop_table) = 2
132
+
133
+
In this scenario, the airspeed sensor was flagged as invalid by the innovation check four seconds after the blockage occurred (shown by the middle plots). This delay consisted of:
134
+
135
+
- 2 seconds to accumulate enough error in the innovation integrator (per ASPD_FS_INTEG)
136
+
- 2 seconds to satisfy the ASPD_FS_T_STOP hold time before the failure was declared
137
+
138
+
Had the [load factor check](#load-factor-check) been enabled, the check would have been triggered immedietly (shown by the right plots).
This example involves a vehicle that experienced pitot tube icing during flight. The configuration was:
145
+
146
+
[ASPD_DO_CHECKS](#aspd_do_checks_table) = 23 ([missing data check](#missing-data-check), [innovation check](#innovation-check) and [first principle check](#first-principle-check) enabled) \
147
+
[FW_PSP_OFF](#fw_psp_off_table) = 4 (degrees) \
148
+
[FW_THR_TRIM](#fw_thr_trim_table) = 0.6 \
149
+
[ASPD_FP_T_WINDOW](#aspd_fp_t_window_table) =2 \
150
+
[ASPD_FS_T_STOP](#aspd_fs_t_stop_table) = 2
151
+
152
+
Pitot icing primarily triggers the first principle check. This check requires four consecutive seconds of invalid airspeed:
153
+
154
+
- 2 seconds of inconsistent behavior (as defined by ASPD_FP_T_WINDOW)
155
+
- 2 seconds to satisfy the failure hold time (ASPD_FS_T_STOP)
156
+
157
+
This time window is shown as a blue box in the graph below. Within it, the vehicle's throttle is more than 5% above trim, the nose is pitched down, but the indicated airspeed is not increasing; indicating a likely blockage.
The innovation check would not have triggered until more than a minute later. This delay occurred because the CAS scale estimate was adapting to compensate for the discrepancy between ground speed and measured airspeed.
162
+
163
+
The top-right graph shows scale adjustments working to "straighten" the validated TAS in the top-middle plot.
164
+
165
+
As a result, the innovation metric stayed within limits for some time. Only when the measured airspeed began decreasing too significantly would the innovation check have failed, shown by the bottom-middle plot.
0 commit comments