GNSS – We are almost there…

It is funny isn’t it? I took me almost another year just to finish mostly LNAV and VNAV paths projection algorithms. So where are we today?

GNSS simulation has today ALL features of real HT1000, except for Inertial position source (but this one is not used at ATR). LNAV path tracking, (bank control to autopilot) also works, but it has not been heavily tested yet. There is only one thing missing – it’s the LNAV path capture mechanism.

This algorithm is used, when aircraft is in more than given (~ 3NM) distance from active leg. It has normally different behaviour based on A/C position and course to the active leg. Such thing is so hard to test on simulators, I’ve seen so far two professional sims, but each was performing slightly different. It is also hard to test in real flight – as you have to first significantly deviate from active RTE to test LNAV capture. To be honest I do not know where to find this information at all. So I will probably have to guess. 🙂

Anyway as the last TODO is LNAV guidance testing I will need to finish AFCS as well to be able command ATR autopilot by GNSS. This should take me 1-2months and I will have some free time in near future. Hopefully I will be able to update you more often from now. 🙂

 

GNSS project update

Here is  regular project status update.

Currently there are two important issues I work on:

• How to command AP to follow LNAV track (i.e. guiding AP to intercept and follow constructed track)
• How to properly construct VNAV descent path.

For LNAV navigation I am currently playing with FSX autopilot commanding it in HDG mode, to follow LNAV track. I am also working on track intercepting logic. The results are promising, although the testing itself is boooooooooring. 🙂 I have logged hours and hours of flight time in FS, just watching if AP works as expected.

Regarding the VNAV I analyse possibility to use Dijkstra graph method to construct descent path. Still there are some blind spots in my knowledge, how GNSS works in some special cases of descent profile, but I’ve already prepared some tests cases to be done in plane. Hopefully the result will help me to deal with VNAV simulation.

When I will finish these two challenging issues, there will be first beta release. Meanwhile, in the time I am too tired by trigonometry functions, I am checking page by page, function by function  GNSS against Honeywell manual and videos made in real ATR and certified simulator. I am still finding some minor discrepancies, but most of them are just cosmetic issues. 🙂

GNSS HT1000 – latest news

Here are some hot news 🙂 about GNSS development:

• brand new font was created – AS REALISTIC as it could be – it was mapped from macro pictures from real GNSS screen pixel by pixel into TrueType font curves.
• whole display/screen routines were reprogrammed to have more realistic look – now display is very real. 🙂
• options to display click-able full panel bitmap was added – usable for example for touch screens.
• there are huge tests ongoing – GNSS simulation is compared against behaviour of real GNSS – more than 120videos was taken in real ATR and certified simulator (~11GB of  data)

Here is short video of new font and panel like interface. Testing machine is old P4 CPU, so video is little bit choppy, because of high load from video capture SW. 🙂

GNSS HT1000 is moving forward

I would like to just briefly inform about project status. Yes It took more than I’ve supposed, but it project is ongoing and it is moving forward – just check what is done and how few things are ahead. 🙂

GNSS HT1000 status 22/07/2011

So here is the current status of HT1000 GNSS simulation

First some video as a prove, that I am not kidding. 🙂

 

And bellow some written status, page by page:

GNSS HT1000 status 01/06/2011

I continue with HT1000 system development.

For today 01/06/2011  most of internal logic on HT1000 program is done.

I have currently:

• procedures to parse and store Navigraph Level-D AIRAC
• working interface to MSFS (read/write)
• WGS84 Ellipsosid model
• GPS satellite position model
• internal structures to capture keys
• subroutines to format pages.

Now I am writing FMS logic page by page.

You can see my first in flight test somewhere over Channel here:

GNSS HT1000 play

Honeywell Trimble HT1000 is GNSS (Global Navigation Satellite System) used in newer ATRs. As I have some time waiting for ELMA E37 encoders order, I’ve start playing with simulation of this system.  As described in post about AFCS, this system will be standalone, running on any network PC, communicating with cockpit HW and simulator via Opencockpits IOCP (SIOC) server. For display this nice LCD 5″6 display will be used.