Bio15 From Colombia, joined Mar 2001, 1089 posts, RR: 7 Posted (10 years 2 weeks 1 day 9 hours ago) and read 2821 times:
Im on a control-systems course right now, and we are designing a project to control a plant we choose. I think it is pretty interesting to model an aircraft's autopilot system, which is basicaly a control unit where the input parameters are the dials in the cockpit. Yet, I need to either find the model or make an approximation of it.
I was thinking of the vertical navigation modes combined with some lateral navigation, nothing very complex. First I would need to know which are the VNAV modes, and how they work basically. Secondly, I need to know what constraints are put by the manufacturer such as vertical acceleration, or elevator deflection, if there are any. I appreciate any information at all on this subject. I know it is a complex subject, but I am only looking into the basics right now. Thanks in advance.
OzLAME From Australia, joined Feb 2005, 338 posts, RR: 0
Reply 1, posted (10 years 2 weeks 22 hours ago) and read 2631 times:
I'll start off by saying I do not work on jets with truly sophisticated autopilots, but I can get you started. With a couple of specific exceptions that are irrelevant to what you are doing, all autopilots that have pitch control have an Altitude Hold mode. This mode usually uses a signal from a baro/static pressure-sensing device such as the Air Data Computer or a dedicated Altitude Capsule (supplied by the A/P manufacturer); the pilot has to first establish the a/c at the desired altitude and then engages the ALT HOLD mode, the autopilot will then maintain the a/c at the specific static pressure at which the mode was selected, usually by using the pitch trim. If the QNH changes the pilot will have to change altitude and then re-select ALT HOLD (this of course does not happen with airliners or other higher-performance a/c that cruise above 11 000' and use 1013.25mB/29.92"Hg as their reference pressure).
If the pilot of an a/c fitted with one of the better A/Ps wants to change the level of the a/c, say to descend from FL180 to FL140, then he/she could dial in 14 000' on the Altitude Preselector/Alerter and the A/P will descend the a/c and level off at FL140. On the a/c I work on, the APA uses the Mode C Grey Code outputs from the Encoding Altimeter and is therefore only able to control the altitude to the nearest 100'.
The next mode all decent muti-axis A/Ps have is APPROACH. This mode is selected/armed prior to commencing the approach to an airport and when the A/P is fed a Valid Glideslope Deviation signal (which is actually two signals, the Deviation Signal itself and a separate Flag signal) from the G/slope receiver, the A/P changes from Armed to Capture and maintains the a/c on the Glideslope.
There is also VERTICAL SPEED Mode, but the A/Ps I am familiar with do not have this mode. AFAIK the a/c will climb or descend at a certain Vertical Speed as selected by the pilot; I think the input signal is a rate-of-change-of static pressure from the Air Data Computer.
The LH photo below shows a King Air instrument panel, you can see the Altitude Preselector/Alerter to the left of the radar display.
Bio15 From Colombia, joined Mar 2001, 1089 posts, RR: 7
Reply 2, posted (10 years 1 week 6 days 23 hours ago) and read 2593 times:
Hi OzLAME, that information is absolutely helpful. What I understand now is that the altitude hold mode takes the APA as the input reference signal, and the output signal is the static port information, I assume amplified to the proper gain. In the case of the approach mode the input would be the glide slope signal, and the output would be a rate of change of static port info (vertical speed). Is that right?
Do you happen to know any mathematical model for this system (transfer function, state variable, or simple mathematical equations) ?
I value your input on this subject, especially since you're into avionics. Thanks again