Feasibility and Product Selection Advice Required For Hardware in the Loop Model

Hi,
I am new to the Arduino ecosystem, and unsure what products to buy to create the system I am aiming for.
I have attached a very rough sketch of the system I am trying to create…
I am dynamically modelling a wind driven car in Simulink. I am powering the car via a motor, which is run by a Arduino. This motor needs to wirelessly connect to the Simulink program on the computer.
I am also producing the resistive forces produced by the wind turbines by attaching a wire to the chasis of the car. The rate at which the wire is released (modelling the resistive forces) is controlled by another motor, run by an Arduino, which is again connected wirelessly to a Simulink program.
I have done quite a lot of research, but I would like some more experienced opinions on component selection and project feasibility.
From what I understand, I need two Arduino Leonardo boards, both with motor and wireless shields, with Xbee wireless transmitters. I also need two external power supplies for the motors (9v batteries…?).
As well as sending the Simulink program information to each Arduino, I also need them to transmit system information back to the computer (such as motor position, current drawn ect…), so this is two way communication.
If possible I would also like to attach a strain gauge (not shown on diagram) between the vehicle chassis and the wire, and send this information through the Arduino on the car, back to the computer.
I would like to know;
A - is this feasible
B - have I chosen the best Arduino board for this design
C - which is the best Xbee module for this design
D - what sort of sampling rate could I expect for this design…
E - what is the best way of connecting the Xbee modules to the computer (is a XBee Explorer Dongle the easiest answer?)
F - what kind of motors should I be looking out to make this possible…
Thank you so much for your time, I am sure these are stupid questions. Please keep in mind the fact that I am relatively new to these systems. Try and keep explinations simple…!! I am sure I will have more questions in the future. Thanks again
xx

Why is it necessary to have wireless links when everything is connected by tension wires anyway? - they can't be all that far apart.

If you are content to use wired connections a single Uno would probably do everything and save a lot of money.

If you must have wireless connections you might wish to consider these Deltino devices that have a 2.4GHz transceiver and a H-bridge for motor drive as well as an Atmega 328 http://www.deltino.com/. I am using them to control model trains. Another possibility is the Moteino which contains a 328 and a transceiver http://lowpowerlab.com/moteino/#specs_moteinor2. I have no experience of Moteinos.

...R

Thanks Robin..

I originally looked at a wired connection, but it would add another factor to the dynamic model, that would be very hard to predict. (This is being externally funded, so accuracy/simplicity is more important than price for the project) Will have a look at what you have suggested. What are the advantages between what you have suggested and the Xbee modules with a wireless shield? Also what motors do you use for your trains?

Really appreciate your advice!!

Please read my response on a suggested approach to design here:

The experiment:

I am dynamically modelling a wind driven car in Simulink. I am powering the car via a motor, which is run by a Arduino. This motor needs to wirelessly connect to the Simulink program on the computer.

The methodology:

I am also producing the resistive forces produced by the wind turbines by attaching a wire to the chasis of the car. The rate at which the wire is released (modelling the resistive forces) is controlled by another motor, run by an Arduino, which is again connected wirelessly to a Simulink program.

Here is where I am having my problem:
I ask myself if the methodology is appropriate; that is, all of the wind forces are being combined into one vector but your attachment of the resulting vector to the model may lead to completely inappropriate resolving forces. Said another way, if you do not model with wind, how can you be sure your aggregate vector is applied correctly to the model? This troubles me but maybe you are not concerned about overly accurate results in this area.

Scope creep:

As well as sending the Simulink program information to each Arduino, I also need them to transmit system information back to the computer (such as motor position, current drawn ect…), so this is two way communication.
If possible I would also like to attach a strain gauge (not shown on diagram) between the vehicle chassis and the wire, and send this information through the Arduino on the car, back to the computer.

So, during conceptual design, you are already planning on enhancements. This is not good. You should list your requirements for this phase and make a second sheet for the next phase - then forget the second phase for the moment. As you craft your 1st phase solution, you should continually address the second phase list to ensure that you are not going down a dead-end road that will preclude requirements in the 2nd phase. Sometimes, you just have to cross-out some of the 2nd phase wants.

You select your hardware based on requirements that have been vetted through the analysis considering the time-to-market (or when you have to complete a project for a grade), your budget, and your 2nd phase list of wants. Realistically, is the 2nd phase going to see the light of day? Are you willing to overspend on the 1st phase to have reserved capacity for adding phase 2? Entire disciplines exist on this thinking methodology - have a look at the PMP methodology.

A - is this feasible
“You can do anything with a computer if you have enough time and enough money.”

B - have I chosen the best Arduino board for this design
“I have not seen a design… just a sketch and want list.”

C - which is the best Xbee module for this design
“See ‘B’ above”

D - what sort of sampling rate could I expect for this design…
“What is your requirement?”

E - what is the best way of connecting the Xbee modules to the computer (is a XBee Explorer Dongle the easiest answer?)
“It is A WAY to connect.”
https://www.sparkfun.com/pages/xbee_guide

F - what kind of motors should I be looking out to make this possible…
My guess: variable reluctance linear motor”
Assuming you need calibrated force resistance in an acceptable rotary torque nomenclature… I guess you could do your own calibration or use a (calibrated) force transducer.

Forums work best when you have ONE (maybe two) very pointed questions. Pumping the forum for a project design is problematic because you will either get NO response or you will get many non-focused responses because no one understands what the heck you are thinking or why you are doing what you are doing. Bad use of everyone’s time.

Ray

Hi Ray,

Thanks for your post. I do get your point, and realise that the question is very open ended, which I understand is frustrating. This is an important project, I am very inexperienced, and I just wanted a more experienced view to check I wasn't attempting something insane/ I was ignoring a far easier way of doing things before I went any further along the line!

Vis a vis the methodology, I understand your point, I have been thinking about the same problem myself, and haven't figured out an easy way around it. It is not as simple as modelling with wind, as there are dynamic mechanical losses in the turbines, as well as the resistive forces produced, plus inefficiencies (there are in fact two turbines, one behind the other, so the air flow isn't even laminar), that are all very hard to model. Purely for simplicity, I think that the "aggregate vector" needs to be applied at a single point, although this will of course create inaccuracies.

I do need to plan in a more structured fashion, thanks for you advice! I can't find any links to PMP methodology, am I being thick?

I guess the most important question is should I get the arduino uno, or mega board. I think it is between these two, for a number of reasons. I would also like to know if it is possible to stack a motor and wireless shield on these boards. (It appears that they both use pins 11 through 13, does this make them incompatable?

Thanks again Alex

@pitters: I am so glad you did not take offense- it is difficult to (for me...) articulate concisely assistance without specifics. Oh, on the PMP stuff: http://www.pmi.org/Knowledge-Center/Knowledge-Shelf/Project-Management-Methodologies.aspx

arduino uno, or mega board Personally, if I'm walking on a slippery slope, I want a safety-net and you get 4x the SRAM on the mega2560 plus more flash. I have used the 2560 to model a multi-sender engine monitor for a Rotax experimental and never had any problems: realtime date-time, 4 engine temps, ampmeter, voltmeter, 2 thermocouples for exhaust gas, full logging diagnostics and 2 OLEDs for cockpit displays.

You may wish to send forum moderator Crossroads a PM about his custom board solutions. I have 3 of his products and am very pleased! http://www.crossroadsfencing.com/BobuinoRev17/

Ray