Nonlinear system identification using neural network (black box model)

Hello, my thesis work is based on "suface EMG- angular acceleration modeling using different system identification techniques"......can anyone help me in doing nonlinear system identification using neural network...

Well, look at that. I actually had this problem before--and SOLVED it before! [facepalm] I'd forgotten all about it....
https://bbs.archlinux.org/viewtopic.php?id=140151
I just added "vmalloc=256" to my linux line, and X started right up!
[edit] Well, mythtv had the solution, as well: http://www.mythtv.org/wiki/Common_Probl … _too_small
Last edited by wilberfan (2012-11-05 19:38:06)

Similar Messages

System identification error 41263

hi...
can somebody please explain why i am getting an error "41263-The System Identification Toolkit cannot retrieve the data in system model because the model type you wire in is incompatible with the system model requirements. " while using model simulation vi to simulate an arx model i created using estimate arx model vi?

Hi ajaleelp,
Sorry for the delayed response on this, but I have a few questions for you. Are you just receiving this error at run-time? Meaning you have no broken wires or anything, but the VI stopsat the model simulation VI?
Also, have you tried our example for model simulation? You can find it here:
C\program files\national instruments\labview [version]\examples\System Identification\Getting
Started\General.llb
This actually uses the estimate arx and model simulation VIs. So, you should be able to see if that works, and then compare it to your system.
Hope this helps!
Chris Van Horn
Applications Engineer

System identification for open loop unstable plant?

Hello,
I have been trying to use system identification on labview. My problem is that my plant is open loop unstable and none of the tutorials or examples I have found seem to cover this so I cannot get it to work. Does anyone have any suggestions as to how I should go about it? My plant has an RC servo so it needs a PWM signal to stimulate it.
Thanks
Adam

adamkse wrote:
I want to control my plant with lead/lag control but I do not know the plant model which makes this very difficult. I cannot theoretically calculate it because the information I need to obtain a numerical transfer function or state space model is unavailable.
I have it controlled with PID at the moment but this is not ideal.
I am trying to use the system identification toolkit to obtain a transfer function but I do not know what to do because the examples do not work for open loop unstable plants.
Adam
Ok Open loop unstable functions cannot be predictably controlled (hence the terms; open loop and unstable,) Not being able to model the actions is a bit of a problem too.
Somehow, someway you need to get some kind of model or some kind of feedback to either provide stability in the plant or close the loop.
Adding hardware (sensors) is the most likely solution.
tell me more about the plant, I do this kind of thing for a living, so I could offer some ideas.
Jeff

System Identification Assistant Signal Generator

Hello,
I am using a GW Instek power source to heat a graphite rod within a vacuum chamber. I am detecting the temperature with a K type thermocouple connected to a thermocouple input module. I have written a labview program with a PID and have everything running, yet i force tuned the PID. I would like to use the signal identification tools labview provides to identify the transfer function and model the process. I have used the system identification assitant in attempting to send a step signal and acquire a response, yet I have been running into difficulties. The only way I can perform a step function is to manually generate an output of a specified power and then measure the response, as the signal generator within the system identification software is not identifying the GW Instek PSW 30-108 power source. With manual input and importing both input and output data i am able to get a parametric model, yet the transfer function provide me PID values which are 5 orders of magnitude greater then the ones i am currently using. Ideally i would like to use the software to generate the signal and measure the response continuously rather then using data i collect through my program. If possible, I would also be interested in perhaps directly integrating VI's to perform signal analysis directly into my process program.
Problems and what I have done:
I have the drivers for the power source and my process program recognizes the device yet system ident assist cannot find the hardware. I have attempted to creat a task in NI-MAX yet the hardware is also non-identifiable.
Port module input works with VISA and this is how my VI is communicating with the power source.
Goals:
I would like to use the system ident assist to generate power ouput and collect temperature input, and generate a process model, so that I may tune my PID more accuratly.
Also, I may be interested in integrating VI's directly into my process program, yet I am new to the LV language...
I have attached my Labview program below. I would like to thank you in advance for your time and look forward to responses.
Cheers
Attachments:
Test1_RT.vi ‏186 KB

Double post, for further Information see http://forums.ni.com/t5/LabVIEW/System-Identification-Assistant-w-o-NI-Signal-Generator/m-p/3014731#...

RLS for system identification: problem in creating a signal + noise input.

i have successfully implemented the RLS algorithm in labview, but i am having alot of trouble generating an input. at first i used a ptbypt sine input and wired it to my FIR Filter (win coeff ptbypt) and to my data queue (ptbypt) to generate U(n). When i plot the adjusted weights they give a perfect Sinc Function.
So, after this i add in a gaussian White noise VI (ptbypt) and add it's output with the output of the sine wave VI (ptbypt) and like before wire the output of the summation to the FIR Filter and the data queue point by point. The problem that is occurring is that my tap weights after becoming constant do not change. they are the same for when i enter the sine wave alone and when i enter the sine wave with the gaussina white noise, tap weights are same. they only change when i change the low cutoff of the FIR FIlter.
its would be very much appreciated if someone may help me solve this problem i am having. I made the RLS (for system identification) as Final Year Engineering Project.

the input is a sine wave , the sine wave passes through a system which incorporates a noise in it (we say the noise incorporated by the system is the Gaussian white noise) the corrupted input enters the system , the filter weights are adjusted by the algo, and a sine waveform is extracted filtering the noise.
what kind of LTI system. Can you please elaborate? and i cant switch to LMS my project is RLS.

System identification miso models

Hi,
I want to know is it possible to identify the MISO and MIMO models using system identification toolkit?
I m using fieldpoint 2010. The program is running very slow inside the fieldpoint. so is it possible to divide the program in such a way that few compuations(which are less critical) run inside the computer and few inside the fieldpoint?
Regards,
Sai

Hi Ruser,
LabVIEW System Identification toolkit provides SI Convert to Models of CDT VI to convert a model you created in the LabVIEW System Identification Toolkit into a transfer function, zero-pole-gain, or state-space model that you can use with the LabVIEW Control Design and Simulation Module.

System identification for power converter

Dear All,
i am working on system identification for the buck converter . for this i generated variable duty cycle with 20KHz switching frequency on analog output AO1 of usb 6221, now i want to introduce perturbation PRBS in steady state. now my problem is to add perturbation on analog output. please suggest me how to do this on analog output.
thanks
with regards
gurpreet singh

Do you want to mix some signal to your PWM signals? You can use normal add function to add your AO PWM signals with the PBRS perturbation signal.

Neural network: is there any toolkit?

Is there any toolkit in order to use neural networks with labview? (I am not an expert about neural networks, I have just been said today to try to solve a problem using neural networks, I even don't know where to start from...well..I am starting from labview!).
Solved!
Go to Solution.

if you want to just use it and have it simple use this one: https://decibel.ni.com/content/docs/DOC-41891
Best regards, Piotr
Certified TestStand Architect
Certified LabVIEW Architect

Strange Black Box around every prompt I am given...

This is very strange. I started up iChat for the first time on MBP and am now having a very strange occurance. At every prompt there is a strange black box around it. For example, right now there is a black box around the back arrow in my safari browser. When I open the finder there is a black box around my hard drive. I just opened system preferences and there is a black box around the back arrow in system preferences box...
Can anybody help me?
MacBook Pro 2GHz, PowerBook G4, PowerMac G4 Mac OS X (10.4.5)

heheh ya it's funny when you look at the people who just hear the mac talking and they don't know wat to do.. but ya.. universal access is only for that.. for the F8 key is exposè, which is nothing related to Universal Access at all, so you can keep Universal Access options off and exposè will still running
Anyway, you can configure both from the System Preferences.

Black boxes w/numbers inside on scanned docs

We are using Adobe Acrobat Standard version 7. When we open scanned documents that were created using Acrobat 5, black boxes with numbers in them appear on the page. Sometimes they appear next to the actual fields that were filled in (i.e. name, date, etc.) but sometimes the boxes and numbers randomly appear on the right-hand side of the page. Any help you can give is greatly appreciated. Thanks!

It simply sounds like the settings you have set in the scanner software is different.
when I scan 20 pages of B/W text I get around 16 MB. With a PC it's around 400 kb.
That's not right.
A single 8.5 x 11 page @ 300 dpi is gonna be ~1.2 MB.
To get 50 pages to be only 400kb, they are being compressed.
You are scanning as a jpeg on the PC and as a bitmap on the Mac.

Using threads in a neural network

Hello,
I've written a neural network and I'm wondering how I could use threads in it's execution to 1) increase (more precisely achieve!) learning speed and 2) print out the current error value for the network so that I can see how it is working without using the de-bugger. Basically, i've read the Concurrency tutorial but I'm having trouble getting my head around how I can apply it to my network (must be one of those days!)
I'll give a brief explanation of how i've implemented the NN to see if anybody can shed any light on how I should proceed (i.e. whether it can be threaded, what parts to thread etc.)
The network consists of classes:
Neuron - stores input values to be put into the network and performs the activation functions (just a mathematical operation)
WeightMatrix - contains random weights in a 2-D array with methods for accessing and changing those weights based on output error
Layer - simply an array that stores a collection of neurons
InputPattern - stores the values in an array and target value of a pattern (e.g. for logical AND i would store in pattern[0] = 1; pattern [1] = 1; target = 1;)
PatternSet - set of InputPatterns stored so that they can be input into the network for learning
NeuralNetwork - the main class that I want to thread. This class contains multiple layers and multiple WeightMatrices (that connects the neurons in each layer). The learn algorithm then uses the methods of the previous classes to generate neuron inputs and ouputs and error values given a specific input. It uses a loop that iterates through as follows:
    public float learn(PatternSet p)
        InputPattern currentPattern = null;
        double netError=0f;
        float previousError=0f;
        float outputValue = 0f;
        float sum=0f;
        float wcv=0f;
        float output1=0f;
        float output2=0f;
        float currentError=0f;
        float multiply=0f;
        float outputError = 0f;
        float weight = 0f;
        int count;
        int setPosition=0;
        int setSize = p.getSetSize();
        Neuron outputNeuron = layers[getNumberOfLayers()-1].getNeuron(0);
        //execute learning loop and repeat until an acceptable error value is obtained
        do
             //set input layer neuron values to pattern values
            currentPattern = p.getPattern(setPosition);
            for (int i=0; i<currentPattern.getPatternSize(); i++)
                layers[0].getNeuron(i).setNeuronInput(currentPattern.getValue(i));
            currentError = layers[getNumberOfLayers()-1].getNeuron(0).getOutputError();
            for (int a=0; a<layers[getNumberOfLayers()-1].getNumberOfNeurons(); a++)
                //set target value of output neuron
                layers[getNumberOfLayers()-1].getNeuron(a).setTarget(currentPattern.getTarget());
            //iterates between weight layers - i.e. there will be a weight matrix between each layer of the NN
            for (int i=0; i<getNumberOfLayers()-1; i++)
                for (int j=0; j<layers[i+1].getNumberOfNeurons(); j++)
                    sum =0f;
                    count=0;
                    for (int k=0; k<layers.getNumberOfNeurons(); k++)
weight = weights[i].getWeight(k,j);
outputValue = layers[i].getNeuron(count).getOutput();
multiply = layers[i].getNeuron(count).getOutput() * (weights[i].getWeight(k,j));
//add values
sum = sum + multiply;
count++;
//check that all weighted neuron outputs have been completed
if (count == layers[i].getNumberOfNeurons())
//pass results to neuron
layers[i+1].getNeuron(j).setNeuronInput(sum);
//activate neuron
layers[i+1].getNeuron(j).neuronActivation();
//calculate output error of neuron for given input
layers[i+1].getNeuron(j).calculateOutputError();
//check that output layer has been reached and all neurons have been summed together
if (i == getNumberOfLayers()-2 && count == layers[i].getNumberOfNeurons())
outputError = layers[i+1].getNeuron(j).getOutputError();
netError = layers[i+1].getNeuron(j).getNetError();
for (int a=getNumberOfLayers()-1; a>0; a--)
for (int b=0; b<layers[a-1].getNumberOfNeurons(); b++)
for (int c=0; c<layers[a].getNumberOfNeurons(); c++)
output1 = layers[a-1].getNeuron(b).getOutput();
output2 = layers[a].getNeuron(c).getOutput();
wcv = learningRate * (outputError) * output1 * output2 * (1-output2);
weights[a-1].changeWeight(wcv, b, c);
learningCycle++;
if (setPosition < setSize-1)
setPosition++;
else
setPosition=0;
while (netError > acceptableError && learningCycle < 1000000000);
return currentError;
}At the moment the net doesn't seem to learn to an acceptable degree of accuracy, so I was looking to use threads to monitor it's error value change while I left it running just to ensure that it is working as intended (which it seems to be based on NetBeans debugger output). For the moment, all I'm aiming for is an output of the netError value of the NN at a particular time - would this be possible given my current implementation?
Thanks for the help,
Nick

For huge NN and a really multi-core CPU (reporting to OS as a multiple CPU's) one may benefit from having:
- an example pump
- a separate threads for calculation of forward and backwards propagation with in/out queues.
Example pump pumps one forward example to each forward processing thread. It waits for them to complete. Then it reads their output and finds errors to backpropagate. It pumps errors to back-propagation threads. They finds weights correction but does not update weight matrix only are pushing them to the output temporary arrays. The example pump takes those corrections, combines them an updates weights.
Redo from start.
The rule of thumb for high-performance is - avoid locks. If must access data which are changing, make a copy of them in bulk operation, prepare bulk result and read-write in bulk operations.
In this example a whole bunch of weight matrixes and states of neurons are such a kind of data. Each thread should use separate copy and the teaching pump should combine them together. This makes one to split data in two blocks - non-changing, common for all threads (the geometry of NN and weights) and changing, separate for each thread (weight correction, in/out of neurons).
Avoid "new", "clone" and etc. for the preference of System.arraycopy on existing data.
Regards,
Tomasz Sztejka.

When accessing shared folder - 'You might not have permission to use this network resource" .. A device attached to the system is not functioning

On a Windows 2008 R2 server that had been working fine, all of a sudden some shared folders became inaccessible. Clicking on them returned the following: "<Folder> is not accessible. You might not have permission to use this network
resource. A device attached to the system is not functioning."
If I create another share with a different name for that same folder, it works fine. If I delete the original share then recreate it with the same name, I get the same error. However, if I right click on the problematic share and select 'map
network drive' that works. So this would not appear to be a permission issue.
I discovered this problem because the path for mapping the home folder as a property of their AD account stopped working.
I have tried most of the common things found on the internet. I've tried accessing via IP, same issue. While I only have a couple 2003 servers, those can access this resource.
At this point, I'm pretty much out of ideas. Any help would be appreciated. I also have some reports of potential issues with some printer mappings too which I will have to investigate in the morning.
If anyone has a solution to this I would be extremely grateful. Thank you.

Hi,
Can you access the shard folder locally? Is a specific server cannot access the shared folder? Please try to Boot your server in Clean Mode to check if some third-party software cause the issue.
How to perform a clean boot in Windows
http://support.microsoft.com/kb/929135
Best Regards,
Mandy
If you have any feedback on our support, please click
here .
We
are trying to better understand customer views on social support experience, so your participation in this
interview project would be greatly appreciated if you have time.
Thanks for helping make community forums a great place.

Cannot use System Pref's/Network -- dialog box won't go stay gone

G5 running Tiger, browsers, Mail and other internet functions (DSL, PPPoE) work fine BUT: if I boot the Network app in System Pref's, it shows a dialog box that says "Your network settings have been changed by another application." If I press OK the box goes away for a second and then returns -- there is no way to proceed, or use this app -- the only thing I can do is for Systems Pref's to quit. In other words, it won't take OK for an answer!
Some notes: 1) Network diagnostics from the Help menu says everything is normal. 2) I did move my Ethernet connection to another computer to run that one -- this Mac was off at the time. I did not change any modem or router settings, and this Mac, as I said, works fine except for the above problem. 3) an app called "Internet Connect" in the apps folder will not boot (don't know what this is); 4) the utilities Netinfo Mgr and Network Utility don't tell my anything useful or at least that I can understand. How do I reset the SystemPref's/Network app or whatever it is so I can use it? Thanks.

Hi,
Check out this thread: file:///Users/carolynssamit/Desktop/Mac%20Troubleshooting/Your%20network%20setti ngs%20have%20been%20changed...%20%20fix.webarchive
And here: http://forums.macrumors.com/showthread.php?referrerid=59552&t=592305
And more here; http://forums.macworld.com/message/666874
Carolyn

TA38556 I am trying to set up my apple tv using the hospital network. I am able to detect the network and chose it. But the way the system works is that you have to agree to the terms and conditions of using their network. How do I accomplish this with ap

I am trying to set up my apple tv using the hospital network. I am able to detect the network and chose it. But the way the system works is that you have to agree to the terms and conditions of using their network before access the internet. How do I accomplish this with apple? Please help.

There are some other options to you, if you're not at home and nobody is using your router you could set it up at the hospital, if this isn't an option you could get a second router and use it. Many ethernet routers are available on the cheap, and ethernet cables are cheap too.
Also, a macbook can share it's internet via the built in wifi it has. See this go to settings-sharing-internet sharing and configure as desired. This may require an internet connection via ethernet. But you should be able to create a wifi network for using something like airplay/home sharing.

Dv9500t system board identification

I need to replace the system of my HP Pavillion dv9500 t laptop. the service manual provides two possible part numbers viz. one for a board "with 256MB of videp system memory for laptops with discrete video system memory" and a second where the 256MB is replaced by 128MB. I have a 512MB Nvidia GeFore8600M GS graphics card. So, which one is right for my system the part number with "256MB" of video system memory" or the the one with "128MB". System information tell me tha the board manufacturer is Quanta, system ID is 30CB and version is 79.2E
This question was solved.
View Solution.

neveo wrote:
I need to replace the system of my HP Pavillion dv9500 t laptop. the service manual provides two possible part numbers viz. one for a board "with 256MB of videp system memory for laptops with discrete video system memory" and a second where the 256MB is replaced by 128MB. I have a 512MB Nvidia GeFore8600M GS graphics card. So, which one is right for my system the part number with "256MB" of video system memory" or the the one with "128MB". System information tell me tha the board manufacturer is Quanta, system ID is 30CB and version is 79.2E
You need the 8P PM Intel system board. It is the
(14) With 256 MB of video system memory for use with computer models
with discrete video system memory HP p/n 447983-001
Best regards,
erico
****Please click on Accept As Solution if a suggestion solves your problem. It helps others facing the same problem to find a solution easily****
2015 Microsoft MVP - Windows Experience Consumer

Nonlinear system identifica​tion using neural network (black box model)

Similar Messages

Maybe you are looking for

Nonlinear system identification using neural network (black box model)