Running SubVIs with state machines in parallel

Hi,
I'm new to labview, working on it for a couple of weeks.
I'm building an application where there are a number of identical elements atht each is controlled by an identical state machine which is independent from all others.
My problem is how to place each state machine in a subvi and running them in parallel.
Each of the subvi has a while loop for the state machine, and the program gets stuck inside never entering another subvi.
Is there an elegant way to run a couple of SubVIs each holding a state mchine in parallel?
I come from the VLSI world so there is no problem implementing this kind of logic in hardware description language but the labview way to do it eludes me...
thanks you very much for any help,
Eyal.

braxat2000 wrote:
Hi,
Thank you everybody for your kind help.
The problem was solved and my application is OK now.
For future reference,
I used the execution reenter subvi to allow multiple instances of it in parallel, and added a global variable that the subvis can change, and the main can read from.
Used semaphores to make sure correct parallel write to common sources.
Cheers,
Eyal.
Great solution- but how to avoid race conditions in the global variable? Globals are notorious for causing headaches, nightmares and nausea so if there is a way to avoid them and enforce a bit of cause- effect its highly desirable. In this case you have clones of a re-entrant vi running in parallel so you CAN'T use a functional global (the most common solution see Ben's famous Action Engine nugget)
Since we can't enforce data dependence we need something else as a cause-effect team..... Change event dependency would be a good substitute. Two methods for getting this type of dependence would be a dynamic event or my favorite the "Named Queue" set the data type of the named queue as a cluster of instance of source and data from your global and the main loop will be able to see EVERY write to the queue AND the source. Moreover the queues datatype can be expanded to offer any information the main vi needs to be a good executive. In fact a new sub vi that "monitors status" of the clones would make the application highly scalable and robust.
Jeff

Similar Messages

HT1338 problems with time machine and parallels desktop, mac osx mountain lion 10.8

have a good day, i have been problems with time machine and parallels desktop, dont work in mac osx 10.8, i upgrate two mac´s (macbook air, imac), in the two machines shows problems, could you help me please.
two machines have good ram memory (macbook air 11.1 2010 early , imac 2009 late).
i am waiting for your help.
thanks.
Jesus.

Take a look here:
http://pondini.org/TM/Troubleshooting.html
Specifically, you'll probably want to try item #A5 on that page, but there's a lot of other info there that may be useful.

How to use elapsed time function with state machine in Lab VIEW

Hello
I've been trying to use state machine with elapsed time function in order to sequentially start and stop my code. The arrangement is to start the code for 1 minute then stop for 5 minutes. I've attached the code, the problem is when I place the elapsed time function out of the while loop it doesn't work, on the other hand when I place it inside the loop it does work but it doesn't give the true signal to move to the next state.
Could you please have a look to my code and help me to solve this issue.
Regards
Rajab
Solved!
Go to Solution.
Attachments:
daq assistance thermocouple(sate machine raj).vi ‏436 KB

Rajab84 wrote:
Thanks apok for your help
even with pressing start it keeps running on wait case
could you please explain the code for me, the use of Boolean crossing, increment , and equal functions
Best Regards
Rajab
OK..I modded the example to stop after 2 cycles. Also recommend taking the free online LabVIEW tutorials.
run vi. case statement goes to "initialize", shift registers are initialized to their constants. goto "wait"
"start"= false, stay in current state. If true, transition to "1 min" case
reset elapsed timer with True from shift register(counter starts at zero)."time has elapsed"=false, stay in current state(1 min). If true, goto "5min" case
reset elapsed timer with True from shift register of previous case(counter starts at zero)."time has elapsed"=false, stay in current state(5 min). If true, goto "1min" case. Also, bool crossing is looking for "true-false" from "5 min" compare function to add cycle count.
Once cycle count reaches 2, stop while loop....
Attachments:
Untitled%202[1].vi ‏42 KB

Relay Testing with State Machine

Hi,
Yesterday I very kindly had some help from "crossrulz" in setting out the following procedure. Today, the National Instruments representative called and said I should implement my programme as a "State Machine", which I have attempted in the programme attached. I was wondering how I might now implement the procedures 3 to 5 in this State Machine format as it seemed very complecated in the brief time we had.
What I am trying to do is test relays, using an automated test procedure.
The procedure takes the following steps:
1. A NI9472 DO card and transistor drive turns the relay on/off via energising the coil (needs the time period to be variable).
2. Counts how may cycles the relay contacts have operated, this will easily be in excess of 100,000 cycles.
3. Every 100 cycles (again variable), the contacts are held closed, the load circuit (the current the relay is switching, which will be 15A at 30V) is disconnected, via another relay and a measurement is taken via a separate four wire method. This consists of a 6V supply being placed across the contacts (switched in again via a relay) with a 1A current flowing through the contacts, the voltage drop across the contacts is then inputted into the NI AI NI9219 card (this is in the region of 10’s of mA) and is directly proportional to the contact resistance, which is what I want to measure.
4. The measurement is taken and stored and the measurement circuit disengaged, the load circuit re-engaged, and another 100 cycles is counted before the process goes back to stage 3.
5. This process goes on until the relay fails and hopefully a large set of contact resistance data is gathered.
I have no real experience with NI Labview.
I have looked at the basic academic tutorials and have come up with the following programme for stage 1, however I can't get the DAQ Assistant to output every cycle
Would this be a very difficult programme to implement for the novice? The academic tutorials seem straight forward but when I try to do someting it seems very difficult.
Hoping someone may be able to help, thanks
Andy
Solved!
Go to Solution.
Attachments:
Relay Test State Machine.vi ‏207 KB
Relay Test State Machine.vi ‏207 KB

Hi James,
I've tried to implement some logic as suggested. I can now get the 1A coil to engauge after the nth cycle and I think it's taking a measurement, but can't seem to get the 15 A coil to disengauge at the same moment, which is crucial else the measurement will be across 16 A rather than just the 1 A. Any ideas? I just want to get it working so we can get started with testing.
I'm going to get my supervisor to book me on a course at Newbury for future projects, it's very difficult to pick up of your own back.
Kind regards,
Andy
(Helps if I attach it)
Attachments:
Relay Test State Machine.vi ‏209 KB

Problem with state machine

Hi! I've got a VI intended to control a storage tank. I've tried to do that by creating a simple state machine that changes the value(on/off) of some valves basing on 2 conditions:
-when a certain temperature reaches a pre-set value AND its derivative is <0, the state must change from on to off,
-when some time passes the state must change from off to on.
Unfortunately I can't get this work as it never changes state (e.g if I start with on always remain on "on" and viceversa).
Could you please tell me what I'm doing wrong? I'm attaching the screenshots of the 2 states.
Thanks!
Solved!
Go to Solution.
Attachments:
screenshot1.jpg ‏302 KB
screenshot2.jpg ‏300 KB

It looks like your shift regsiter might not be wired correctly.
Attachments:
screenshot1.jpg ‏303 KB

Help with State Machines

I am implementing my application using state machines, and I have the following problem.
1. I have an initialized state that will initialize my controls.
2. Next I want the next state to check for user input.(boolean button will be pressed to determine which state is next).
3. In one of the states, I want the user to be able to change the initialized values, and afterwards go back to the user input state, to determine which state is next
I have implemented the initialize state as well as the user input state(use event structure for that state), but I do not know how to let the user be able to change the values of my control , then the user acknowledges the change, and the state goes back to the user input state waiting for the next state.
Thanks for any help
Solved!
Go to Solution.

Thank you Ravens Fan, I did not really understand your message. I have a simplified version of what I am trying to accomplish.
1. I have 3 controls and one indicator
2. When the Vi is run , the controls should initialize to the values in the intialize state. The next state will now be the user input state.
3. When at the user input, the user can either press the "analyze button" and the result will be displayed in the indicator. or the the stop button to stop the simulation, or
4. The user can change the values of the control, after all the controls have been changed to his desired values, he can then press "accept values" and the state goes back to the "user input".
Thank you for any help
Attachments:
example.vi ‏26 KB
example control.ctl ‏5 KB

How to interrupt running subvi with global variation

Hi,all
   I want to use global variation STOP to interrupt the running VI at anytime, but the subVI must be running until executing the WHILE. I knew when one action is operated in the EVENT case, the other action will not be iperated. My question is how to solve the problem without POP subVI and the DATA transmitted to MAIN panel. Give me a suggestion, THKS!
Lioncsu
Attachments:
Main VI.jpg ‏43 KB
Sub VI.jpg ‏41 KB

Hi Lioncsu,
      In Main, when a Start event occurs, the Value of Stop (False?) is immediately written to the Global (once), and SubVI starts running.  The Start event case won't complete - or write another value to the Global - until the SubVI is done. I haven't run this, but I think pressing Stop may cause the SubVI to run (again) - for one iteration!
Attached is a simple VI which uses two loops - one to handle GUI events (the buttons) and one to start a while loop.  Maybe you can replace the while-loop with your SubVI?
Hope it helps!
"Inside every large program is a small program struggling to get out." (attributed to Tony Hoare)
Attachments:
Notifiers.vi ‏51 KB

How can I establish an enum typedef for use with state machines and action engines?

Hey all--
I have an action engine with several instances. Any time that I add a method to the engine, I have to modify all of the calling code to update the constant that determines the method being called. There must be a smarter way to address this problem but my LV experience is insufficient to identify it. Can any of you wiz-bangs tell me if it is possible to make a type def (I can do this) and use it with each instance of the action engine call so that changing any instance also updates the typedef?
Thanks.

Take your enumerator. If it is not a control (ie constant), change it to a control.
Under File Menu, select New > Custom Control.
Move (or copy) your Enum Control to the Custom Control. Change the Control to Strict Type Def (the selection to the right of the wrench). Save it.
(EDIT: ** OOps, the picture shows the wrong one.. sorry**)
Copy that new Strict Type Def control to your code.
See images below.
OR.. If you're using LV8.x, and using a project, I would recomment that you do it from within your project explorer. Simply insert a new "Control". The rest of the steps are similar, except that you can insert it from your project.
R
Message Edited by JoeLabView on 07-08-2008 06:47 PM
Attachments:
TypeDef.PNG ‏33 KB

Why can't snow leopard be run as virtual machine in Parallels?

I'm wondering if anyone can tell me why Apple's EULA won't allow Snow Leopard to be run as a virtual machine in Parallels? I am using OS X Mavericks and the latest version of Parallels 8. According to Parallels, in support article ID 112323, the reason it will not work is that it is prohibited to run by Apple's EULA.

Rev. Barry wrote:
I'm wondering if anyone can tell me why Apple's EULA won't allow Snow Leopard to be run as a virtual machine in Parallels? I am using OS X Mavericks and the latest version of Parallels 8. According to Parallels, in support article ID 112323, the reason it will not work is that it is prohibited to run by Apple's EULA.
It was a common Urban Myth that the Snow Leopard EULA prohibited its virtualization in Lion, Mt. Lion and Mavericks on a Mac. That myth has been debunked over the last few years and the remaining debate, if any, has been largely rendered moot by Apple's decision to reduce the price of Snow Leopard Server by 95% to the same $20 charged for Snow Leopard client.
Eric's link will get you to an article that describes how to install Snow Leopard client into Parallels 8, but at the end of the day, it is easier to both install and maintain Snow Leopard Server into Parallels 8, with virtually the same result you desire: the ability to run most PowerPC apps in Mavericks.
Here is the link to Installing Snow Leopard Server into Parallels 8 for DUMMIES:
http://forums.macrumors.com/showpost.php?p=17285039&postcount=564
Rev. Barry wrote:
Are you currently using snow leopard server as a virtual machine? I want to run a power pc app on Mavericks.
Yes, I have experience using Snow Leopard Server in Parallels to run PowerPC apps and I remove the Server apps from the Dock, so as to not be confused by them.
What PowerPC apps do you wish to continue to run?

Lion 10.7.4 issues with Virtual Machines

Hi,
I'm having lot of issues with virtual machines (both Parallels and VMWare)
since having update to lion 10.7.4.
I make myself:
If I switch on Mac OS a run any vitrual machine software
I can run Windows XP and Ubuntu at same time with no issue and very fluently.
But, if I switch on Mac OS and start using Eclipse, Safari 5.2 and other production stuff and
want to use vmware or paralles
they become unusable even if I quit all the tasks ans "purge" memory.
(eg: if i click start button on windows I have to wait seconds before the start menu pops up)
Any Idea?
I have MacBook Pro 13 Unibody mid 2009 with SSD Drive AND 4GB of RAM
and I had no such issues with Lion 10.7.3

alex_75 wrote:
As you wrote before, virtual mchines need big chunks of contigous RAM and
2,5 GB of free RAM wuold have enough ammount of such a RAM.
No, it wouldn't. 2.5 GB of free RAM could consist of 5 chunks of 500 MB of contiguous RAM, separated by RAM allocated by other processes.
Here is a hypothetical memory map. Assume a is application memory, o is operating system memory, and - is free memory.
boot:           oooo--------------
run app1:    ooooaao-----------
run app2:    ooooaaoaaaaaaoo---quit app1:   oooo---aaaaaaoo---
Maybe there is some issue in the way 10.7.4, in some perticolar situations, handle the process of freeing RAM.
What you can't accept, and I don't know why, is that before the update to 10.7.4 everithing was fine.
I don't know why you don't beleve me.
There is nothing in the 10.7.4 update description that mentions a re-write of the virtual memory system. http://support.apple.com/kb/HT5167
I just worte here to see if someone could give any advice to do further invesigations.
I know the price of upgrading RAM but I think that this would be a "microsoft like" solution
and I'm trying to find what service or task or whatever is creating this issue.
The only thing "Microsoft like" about buying more RAM is the fact that you need it to run your Microsoft operating systems. It has been this way as long as I can remember. I upgraded my MacBook to 4 GB RAM in 2008 just so I could run Windows more easily.
Perhaps you have your virtual memory settings too high. Depending on what you are doing in the VM, you may not need the default settings. XP will run low-intensity tasks with a 500 MB VM size. It will run just about anything in 1 GB. Linux should run fine in the same range too.

State machine VS producer consumer architecture - Time Analysis

After learning various methods to program efficiently and learn how to use SM and Producer consumer. I built a program to control stepper motor in both these techniques.
Here is the RESULT. As we can see a simple state machine without any complicated producer consumer technique performs faster than the second one.
I am not sure which is still the best based on performance and optimization. Please advise which one should I keep and why.
Abhilash S Nair
Research Assistant @ Photonic Devices and Systems lab
[ LabView professional Development System - Version 11.0 - 32-bit ]
LabView Gear:
1. NI PXI-7951R & NI 5761
2. The Imaging Source USB 3.0 monochrome camera with trigger : DMK 23UM021
OPERATING SYSTEM - [ MS windows 7 Home Premium 64-bit SP-1 ]
CPU - [Intel Core i7-2600 CPU @ 3.40Ghz ]
MEMORY - [ 16.0 GB RAM ]
GPU - [ NVIDIA GeForce GT 530 ]

This is with state machine alone
Abhilash S Nair
Research Assistant @ Photonic Devices and Systems lab
[ LabView professional Development System - Version 11.0 - 32-bit ]
LabView Gear:
1. NI PXI-7951R & NI 5761
2. The Imaging Source USB 3.0 monochrome camera with trigger : DMK 23UM021
OPERATING SYSTEM - [ MS windows 7 Home Premium 64-bit SP-1 ]
CPU - [Intel Core i7-2600 CPU @ 3.40Ghz ]
MEMORY - [ 16.0 GB RAM ]
GPU - [ NVIDIA GeForce GT 530 ]
Attachments:
Control 1.ctl ‏6 KB
Control 2.ctl ‏6 KB
Motor-UNI_Directiona Dev 3-TIME_ANALYSIS-2.vi ‏54 KB

State machines caveat

I have been using the state machine design pattern in programs for a
while. In this technique, it is a common practice to define boolean
switches to trigger actions in the next iteration of the machine (eg.
"TurnLaserON", "Quit", "Compute", "SetAlarm1"...). Thus, the boundary
between data and processing is easily trespassed.
This sounds to me like a possible source of errors, that should be
limited by some design rule of thumb. I tend to believe that boolean
triggers in state machines should be looked at with the same precautions
as a "GOTO" control structure.
Do you have any experience or theoretical frame to enlighten these
considerations ?
oz

Hi,
I always prefer a state machine with a state buffer. This means, every state
can, potentially, call several other states. In practice, there is a shift
register with an array of strings in it. The first string is called, and
removed. Every state can modify the buffer, so:
+ states can be put before, after or between the buffer
+ states can be removed
+ states can be filtered, sorted etc.
+ buffered states can be copied
This makes programming with state machines a lot easier. You can simply make
a state TurnLaserON, and in the calling state(s) diside what the state after
TurnLaserON has to be. This makes the program a lot clearer.
Some design rules I use (mainlly because if I don't use them, there is no
way of documenting the program):
+ states may only put states before the buffer (I call them sub states)
+ buffered states cannot be removed in a state
Using these rules, I can put the entire program on paper, without any
problem.
Still, I have to transfere information between states. This 'information'
can be a shift register with a bundle, indicators on the front panel (not
visible to the end user), a GLI, buffers, ini file reference, or whatever. I
have never found any big advantage or disadvantage for eny of them, but I
prefer indicators (mixed with buffers when convenient), because they are the
fastest to use (for me), and easiest to debug (IMO).
Regards,
Wiebe.
"O. Zimmermann" wrote in message
news:[email protected]...
>
> I have been using the state machine design pattern in programs for a
> while. In this technique, it is a common practice to define boolean
> switches to trigger actions in the next iteration of the machine (eg.
> "TurnLaserON", "Quit", "Compute", "SetAlarm1"...). Thus, the boundary
> between data and processing is easily trespassed.
>
> This sounds to me like a possible source of errors, that should be
> limited by some design rule of thumb. I tend to believe that boolean
> triggers in state machines should be looked at with the same precautions
> as a "GOTO" control structure.
>
> Do you have any experience or theoretical frame to enlighten these
> considerations ?
>
>
> oz
>

Allow state machine while loop to end only in 1 of the states?

What is the best way to ensure that the while loop that surronds state machine case structure only ends in a certain one of the states?

I have solved the problem of having the VI not stopping (apparently it was due to an event structure). However, now I have another slight problem. I need the state machine to terminate on either of two conditions:
1) If a stop button is pressed
OR
2) if the end of the sequence is reached.
As I have implemented it so far the VI terminates because of the second condition and does not allow me to terminate it by pressing a 'Stop' boolean button. I know that this might be a bit fundamental and simple but I'm new with state machines on labview.
Thanks.
Attachments:
State machine.vi ‏21 KB

Running subVI in parallel with itself using producer-consumer pattern

I'm using the procuder-consumer pattern to create what I call threads in response to various events much like in Producer Consumer with Event Structure and Queues
I have several physical instruments, and would like to run the exact
same test in parallel on each of the instruments. I have a subVI
that takes as inputs a VISA resource and a few control references among
other things and then performs the desired experiment. I have a
separate consumer loop for each physical instrument, with this subVI
inside each consumer loop.
My test VI worked great; each consumer loop was a simple while loop
incrementing a numeric indicator (not using my real subVI above).
However, my real program can only run one consumer loop at a time much
to my dismay. Reworking my simple test VI to use a subVI to
increment the indicator rather than just explicitly coding everything
resulted in the same problem: only a single consumer loop ran at a time
(and as I stopped the running loop, another would get a chance to
begin). The subVI in this case was extremely
simple taking only a ref to the indicator and a ref to a boolean to
stop, and incrementing the indicator in a while-loop.
Is there a way around this? Now that I've spent the time making a
nice subVI to do the entire experiment on on the physical instrument, I
thought it would be fairly trivial to extend this to control multiple
instruments performing the same experiment in parallel. It seems
only one instance of a given subVI may run at one time. Is this
true? If it is indeed true, what other options do I have? I
have little desire to recode my subVI to manually handle multiple
physical instruments; this would also result in a loss of functionality
as all parallel experiments would run more or less in lock step without
much more complexity.
Thank you.

You need to make your subvi reentrant. Then it can run several instances at any time with each instance occupying its own unique memory space. Click on File - VI Properties - Execution, and check the reentry execution checkbox. Then save the vi.
- tbob
Inventor of the WORM Global

Shutting down a queued event state machine with multiple parallel loops

I am trying to find the best way to shutdown a program that consists of a queued event state machine architecture with multiple parallel loops. Can anyone recommend the best way to accomplish this in my attached VI? (From browsing the forum, this seems to be a common question but I haven't found a solution that works for me.)
I welcome any other feedback on my code as well, if anyone is willing to offer it.
My Program Requirements:
If the user presses the "Shutdown" button, the program should prompt the user with "Are you sure you want to stop the program?" and then either return to the Idle state or proceed to stop the program. Additionally if there is an error, the program should prompt the user with "Clear error and continue, or Stop program?" Then either return to the Idle State or proceed to stop the program.
Details of architecture:
The program consists of 3 parallel loops: (1) an Event Handling loop that enqueues various States to a State Queue, (2) a State Machine which enters the latest state that is dequeued from the State Queue, and (3) an Error/Shutdown handling loop which processes any errors in the Error queue.
During normal Shutdown, in the Event Handling loop the "Program.Shutdown" event case executes, and the States "Shutdown" and "Idle" are added to the State Queue. In the State Machine, the "Shutdown" state is invoked. A special error code "5000" is added to the Error Queue. In the Error/Shutdown handling loop, the "5000" error triggers a prompt that asks the user if they want to stop the program. If the user chooses not to stop, a notifier StopNotif is sent to the "Shutdown" state and "Program.Shutdown" event case with the notification "Go". If the user chooses to stop, the notifier sends the notification "Stop". The Event handling loop and State Machine are terminated if they receive the notification "Stop".
In case of error, the program behaves similarly: if the user chooses to clear the error and continue, the program returns to the "Idle" state.
HOWEVER - if the user chooses to stop the program, the program stalls. The notifier that is sent to stop the Event Handling Loop and State Machine cannot be read because the Program.Shutdown event case and the Shutdown state (which contain the "Wait for Notifier" function) are not active.
I have been able to activate the Shutdown state by enqueuing it in the Error/Shutdown handling loop. But I don't know how to activate the "Program.Shutdown" event programmatically, and thereby access the "Wait for Notifier" function inside it.
I tried to place the "Wait for Notifier" function outside the event structure, but then the Event Handling loop never completes. Placing timeouts on the "Wait for Notifier" and on the Event structure makes the program work, but I want to avoid using timeouts because I don't want to turn my event-driven program into a polling-based program. I'd also like to avoid using variables or property nodes to stop the loops, because that requires creating a control/indicator for something that the user doesnt need to interact with.
Thank you!
Solved!
Go to Solution.
Attachments:
Queued event state machine parallel loops empty.vi ‏46 KB

Thanks for clarifying that, Ravens Fan.
I marked crossrulz as the solution for pointing out User Events to me.
I also adopted Ravens Fan's suggestion to keep the Shutdown procedure out of the Error Handling Loop. This has simplified the architecture by eliminating the need for Notifiers. Instead, I have used booleans in case structures to stop the Event Loop and State Machine, and Release Queue to stop the Error Handling Loop.
For reference, I'm attaching my corrected code.
Thank you to everyone who helped!
Attachments:
Queued event state machine parallel loops empty in progress.vi ‏44 KB

Running SubVIs with state machines in parallel

Similar Messages

Maybe you are looking for